Arabic and Semitic Linguistics Contextualized: A Festschrift for Jan Retso [Multilingual ed.] 3447104228, 9783447104227

Citation preview

Arabic and Semitic Linguistics Contextualized

Arabic and Semitic Linguistics Contextualized A Festschrift for Jan Retsö Edited by Lutz Edzard

2015

Harrassowitz Verlag · Wiesbaden

Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.dnb.de abrufbar. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the internet at http://dnb.dnb.de.

For further information about our publishing program consult our website http://www.harrassowitz-verlag.de © Otto Harrassowitz GmbH & Co. KG, Wiesbaden 2015 This work, including all of its parts, is protected by copyright. Any use beyond the limits of copyright law without the permission of the publisher is forbidden and subject to penalty. This applies particularly to reproductions, translations, microfilms and storage and processing in electronic systems. Printed on permanent/durable paper. Printing and binding: Hubert & Co., Göttingen Printed in Germany ISBN 978-3-447-10422-7 e-ISBN PDF 978-3-447-19423-5

 

Table of Contents

LUTZ EDZARD Dedication…………………………………………………………………………………

9

SVEN-OLOF DAHLGREN, ROSMARI LILLAS-SCHUIL, HELENE KAMMENSJÖ, PERNILLA MYRNE, LINA PETERSSON, ANTONIUS VAN REISEN, SINA TEZEL A tribute to Professor Jan Retsö by his pupils…………………………………………

11

LUTZ EDZARD Bibliography of Jan Retsö…………………………………………………………………

13

1. Slavic linguistics SILJE SUSANNE ALVESTAD, Oslo Canonical and non-canonical uses of the imperative in Slavic………………………

19

ANTOANETA GRANBERG, Göteborg New manuscript fragment of a Prolog, discovered in the University Library in Uppsala ………………………………………

40

2. Arabic linguistics and philology WERNER ARNOLD, Heidelberg Living together with the Jews: A Palestinian Arabic text from Jaffa…………………

54

RUDOLF DE JONG, Cairo Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sinai…………

61

WERNER DIEM, Köln Ein Begleitbrief von 904 H zu Erlassen aus dem mamlūkischen Ägypten …………

84

MELANIE HANITSCH, Erlangen ”Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen. Versuch einer Typisierung ………………………………………………………………

102

BARRY HESELWOOD and JANET WATSON, Leeds The Arabic definite article: A synchronic and historical perspective ………………

157

 

6

Table of Contents

OTTO JASTROW, Tallinn The position of Mardin Arabic in the Mesopotamian-Levantine dialect continuum …………………………………

177

ABLAHAD LAHDO, Uppsala Tillo. Two texts reflecting daily life and cultural aspects of the Arabs of Tillo, South-eastern Turkey ……………………………………………………………………

190

PIERRE LARCHER, Aix-Marseille Une occurrence ancienne de la structure kāna sa-yafʿalu en arabe écrit ……………

198

GUNVOR MEJDELL, Oslo Luġat al-ʾumm and al-luġa al-ʾumm – the ’mother tongue’ in the Arabic context………………………………………………………………………

214

MARIA PERSSON, Lund Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic …………

227

ORI SHACHMON, Jerusalem Agglutinated verb forms in the Northern province of Yemen ………………………

260

3. Arabic literature, science, and history of ideas LENA AMBJÖRN, Lund Qusṭā b. Lūqā. On protection against rheum and catarrhs that occur in the winter. English translation of the Arabic text preserved in the manuscript Ayasofya 372………………………………………………………………………………

274

STEPHAN GUTH, Oslo Aesthetics of generosity – generous aesthetics. On the cultural encoding of an Arab ‘national virtue’ ………………………………

299

PERNILLA MYRNE, Göteborg Who was Ḥubbā al-Madīniyya?…………………………………………………………

328

GEORGES TAMER, Erlangen Überlegungen zur Erinnerung und Identitätsbildung im Koran……………………

345

4. Hebrew linguistics SILJE SUSANNE ALVESTAD, Oslo, and LUTZ EDZARD, Erlangen/Oslo Aspect in the Slavic and the Biblical Hebrew imperative ……………………………

 

361

Table of Contents

7

MATS ESKHULT, Uppsala Thoughts on Biblical Hebrew ʾašær ……………………………………………………

391

STEVEN E. FASSBERG, Jerusalem Linguistic variation and textual emendation: The case of Judges 4:20 ………………

397

BO ISAKSSON, Uppsala ‘Subordination’: Some reflections on Matthiessen and Thompson’s article “The structure of discourse and ‘subordination’” and its bearing on the idea of circumstantial clause in Arabic and Hebrew …………………………

405

NAʿAMA PAT-EL, Austin, Texas A note on segholate adjectives in Biblical Hebrew……………………………………

426

OFRA TIROSH-BECKER, Jerusalem Eliʿezer Ben-Yehuda and Algerian Jews: Relationship and language………………

430

5. Aramaic, Ethiopic and comparative Semitic linguistics REBECCA HASSELBACH-ANDEE, Chicago Explicit performative utterances in Semitic……………………………………………

448

JANNE BONDI JOHANNESSEN, Oslo, and LUTZ EDZARD, Erlangen/Oslo Coordinated clause structures in Scandinavian and Semitic involving a finite verb form and an infinitive …………………………………………

486

GEOFFREY KHAN, Cambridge Causative constructions in Neo-Aramaic (Christian Urmi dialect) …………………

506

FEKEDE MENUTA, Hawassa, and RONNY MEYER, Addis Ababa Sonorant alternations in Muher …………………………………………………………

531

SINA TEZEL, Uppsala Arabic or Ṣūrayt/Ṭūrōyo …………………………………………………………………

554

KJELL MAGNE YRI, Oslo Amharic and NSM (Natural Semantic Metalanguage) ………………………………

569

Dedication Lutz Edzard

When Rosmari Lillas-Schuil, senior lecturer at Gothenburg University and one of the former doctoral students of Jan Retsö, asked me on behalf of those in late 2013 to edit a Festschrift for the latter I considered this proposal to be a great privilege and honor, but also a considerable challenge. Indeed, ever since, and even before I came to Norway in 2002 to take on a professorship in Semitic linguistics, this great Norwegian scholar has been an awe-inspiring role model for me. Jan Retsö has an impeccable active command of the ancient and modern Semitic and other languages (especially Slavic and Germanic) he deals with. He also has a virtually unique capacity to understand that certain linguistic phenomena allow, and indeed call for, a multifaceted variety of analyses, as opposed to a stifling monocausal approach so often found in (Middle Eastern) linguistic scholarship. Few scholars have the ability or inclination to communicate in an academic and private context auf Augenhöhe in the way that Jan Retsö has done and continues to do. Jan Retsö has enabled many of us to appreciate the benefits of a fresh scholarly approach that does not automatically take ”Altehrwürdiges” in the discipline(s) for granted. His rich bibliography, encompassing linguistic and philological studies, as well as studies grounded in the history of religion and culture, bears ample testimony of this circumstance. The title of this Festschrift was inspired by Jan Retsö’s equally unique ability to contextualize his scholarship. As a scholar with additional in-depth training in history, he has always preferred to embed linguistic and philological analysis in a historical and realienkundlich framework, be it in the Arab world, in Israel, or in Ethiopia. On the occasion of several visits in Jerusalem, I had the good luck to meet Jan Retsö in the Swedish Theological Institute and to profit from his vast knowledge of languages, religion(s), history, and archeology. His ability to lead historic and archeological tours at places like Jerusalem is unsurpassed, as many colleagues in the field can confirm. Beyond his broad subjectrelated knowledge, Jan Retsö is a humanist in every sense of the word. As much

10

Dedication

as he displays a no-nonsense attitude to academic work, also in social context, he has a broad interest in the belles-lettres, in music, and the arts in general. Last but not least, he has a fine sense of humor. The contributions to this Festschrift by colleagues, pupils, and friends alike reflect many of Jan Retsö’s research interests. I am most grateful to all of my colleagues for having delivered substantial papers that make this Festschrift not only a useful collection of articles but also a true mirror of the state of the art in Arabic and Semitic linguistics, as well as neighboring disciplines. In addition, I owe my colleagues at Gothenburg University, lecturers Rosmari Lillas-Schuil and Pernilla Myrne, many thanks for their inspiration, logistical help and advice. Asbjørn Brovold Gabrielsen of IKOS, University of Oslo, and Kent Hillard helped with various technical matters. Dr. Barbara Krauß and Michael Fröhlich of Harrassowitz-Verlag have been most supportive of the project from the very beginning. IKOS, University of Oslo, generously subsidized the printing of the Festschrift, which the editor gratefully acknowledges. All of us are aware of the fact that this Festschrift meets Jan Retsö at the height of his powers and that everything stated here cannot be of a final nature. This also holds for the attached bibliography, which will certainly continue to grow in the future. For all this we are all deeply indebted to him. Tusen takk, Jan, og lykke til videre ! Erlangen and Oslo, April 2015

Lutz Edzard

A tribute to Professor Jan Retsö by his pupils Sven-Olof Dahlgren, Helene Kammensjö, Rosmari Lillas-Schuil, Pernilla Myrne, Lina Petersson, Antonius van Reisen, and Sina Tezel

We were privileged to be graduate students at Gothenburg University where Jan Retsö was our teacher and Doktorvater, Some of us studied at Uppsala University, but were fortunate to have him as our advisor. His depth of knowledge can be awe-inspiring and like all students who took his courses, we were stunned by this professor who seemed to know everything about the Arabic and Hebrew languages, classical Arabic poetry and the Qurʾān. He belonged, in a good sense, to the old-school teachers, with high expectations and demands. As postgraduate students, we were to spend a lot of time together outside the class-room and we soon found out that this awe-inspiring professor is a genuinely sociable and tolerant person. Although we have become good friends over the years, he has never ceased to be our mentor, generous with his knowledge and always available for questions and discussions. As a scholar, Jan Retsö is a man of integrity, who refuses to jump on trends in the academic world. He taught us to think critically, challenging old as well as new truths. As a mentor, he has always been forthright, gently pushing his students to do their best, but without interfering too much. Besides, Jan Retsö is a man of routine. As long as we have known him, he has had the same lunch at the same time of the day – Swedish crisp bread and Norwegian brown cheese. He gives the impression of being a solid and stable person, who has worn the same kind of clothes and the same hairstyle through the decades. If this made us think of him as a predictable person, we soon found out that he is far from that. He has always surprised us with his vigorous thoughts, and his thought-provoking twists on contemporary as well as historical issues. A discussion with Jan Retsö is typically edifying and inspiring, with unforeseen outcome. Although we have known Jan Retsö for many years now, we are still often astonished by the breadth of his knowledge, which truly qualifies as outstanding. He is a humanistic scholar in the best sense of the word and a polymath; or how could we otherwise describe a man who has knowledge,

12

A tribute to Professor Jan Retsö by his pupils

often in-depth, about ancient and modern Middle Eastern and European history; Islam, Christianity and Judaism and their holy books; European and Middle Eastern literature; classical music, and not least the morphology, syntax, phonology and poetics of languages such as Arabic (including all dialects), Hebrew, Aramaic, Geʿez, Amharic, Sabaean, Akkadian, Persian, Russian, Old Church Slavonic, Polish, Latin, Ancient Greek, Old and Modern English, Old and Modern French, German, Italian, Old Norse, and its modern variants. In addition to his learning, Jan Retsö has an adventurous side, which may be unexpected but comes as no surprise for those who know about his vast travels in younger years. We have to confess, however, that we have often been anxious when he has been away, searching manuscripts in the Ethiopian Highlands, with a lightweight ladder under one arm and a rope under the other. Our concerns are needless, however. When his travel companions fall like skittles due to various ailments, this wise man stays healthy with the help of a little Gammel Dansk. Not to forget his humorous side; Jan Retsö has a keen eye for the hilarious. We have always been looking forward to his return and to listen to interesting and amusing anecdotes from his travels. The bureaucratic thicket of university managing has no supporter in Jan Retsö, and he is an eloquent critic of the control system that has gained ground in recent time. Therefore, we are truly glad that he now will have time to do full-time research. We are looking forward to reading his publications and listening to him present his results at future conferences.

Bibliography of Jan Retsö Prepared by Lutz Edzard

1. Monographs and edited books Retsö, Jan. 1983. The Finite Passive Voice in Modern Arabic Dialects (Orientalia Gothoburgensia 7). Gothenburg: University Press. Retsö, Jan. 1989. Diathesis in the Semitic Languages: A Comparative Morphological Study. Studies in Semitic Languages and Linguistics. Leiden: Brill. Retsö, Jan. 2003. The Arabs in Antiquity. Their History form the Assyrians to the Umayyads. London/New York: Routledge/Curzon. Edzard, Lutz and Jan Retsö (eds). 2005. Current Issues in the Analysis of Semitic Grammar and Lexicon I. Oslo-Gothenburg Cooperation 3rd–5th June 2004. Wiesbaden: Harrassowitz. Edzard, Lutz and Jan Retsö (eds). 2006. Current Issues in the Analysis of Semitic Grammar and Lexicon II. Oslo-Gothenburg Cooperation 4th–5th November 2005. Wiesbaden: Harrassowitz. Watson, Janet and Jan Retsö (eds.). 2009. Relative Clauses and Genitive Constructions in Semitic (Journal of Semitic Studies Supplement Series 25). Manchester: Oxford University Press.

2. Articles and book chapters Retsö, Jan. 1978. “Tefillat Yisrael. Den judisk bönen och dess historia.” Religion och Bibel (Nathan Söderbloms-Sällskapets Årsbok 1978) 37: 64–78. Retsö, Jan. 1979. “Tempelplatsen i Jerusalem i islamsk tradition.” Religion och Bibel (Nathan Söderbloms-Sällskapets Årsbok 1979) 38: 41–53. Retsö, Jan. 1982–1983. “Subjectless Sentences in Arabic Dialects.” Orientalia Suecana 31– 32: 71–91. Retsö, Jan. 1984. “Middle Arabic in a Coptic-Arabic manuscript in the “Röhsska Konstslöjdmuséet”, Göteborg, Sweden.” Studia Orientalia 55: 317–337. Retsö, Jan. 1984–1986. “State, determination and definiteness. A reconsideration.” Orientalia Suecana 33-35: 341–346. Retsö, Jan. 1985. “Commentaries on Islamic conceptions of development.” Queries about Development. A Conference 10–14 September 1984 at the Department of Theory of Science, University of Göteborg, ed. Sven Andersson, 43–48. University of Göteborg: Department of Theory of Science.

14

Bibliography of Jan Retsö

Retsö, Jan. 1987. “Copula and double pronominal objects in some Semitic languages“, Zeitschrift der Deutschen Morgenländischen Gesellschaft 137/2: 219–245. Retsö, Jan. 1987. Pronominal suffixes with -n(n)- in Arabic dialects and other Semitic languages.” Zeitschrift für Arabische Linguistik 18: 77–94. Retsö, Jan. 1989–1990. “The earliest Arabs.” Orientalia Suecana 38–39: 131–139. Retsö, Jan. 1990. “Xenophon in Arabia.” Greek and Latin Studies in Memory of Cajus Fabricius (Studia Graeca et Latina Gothoburgensis 6), ed. Sven-Tage Teodorsson: 122–133. Göteborg: Acta Universitatis Gothoburgensis. Retsö, Jan. 1990–1991. “The domestication of the camel and the establishment of the Frankincense Road from South Arabia.” Orientalia Suecana 40–41: 187–219. Retsö, Jan. 1992. “The road to Yarmuk: the Arabs and the fall of the Roman power in the Middle East.” Aspects of Late Antiquity and Early Byzantium. Papers Read at the Swedish Research Institute in Istanbul 31 May – 5 June 1992 (Transactions of the Swedish Research Institute in Istanbul 4), ed. Lennart Rydén and Jan Olof Rosenqvist, 31–41. Stockholm: Swedish Research Institute in Istanbul. Retsö, Jan. 1994. “The treatment of final syllables in Classical Arabic metres: the linguistic background.” Arabic Prosody and Its Application in Muslim Poetry. Colloquium at the Swedish Research Institute in Istanbul 23-26 April 1992 (Transactions of the Swedish Research Institute in Istanbul 5), ed. Lars Johanson and Bo Utas, 99–106. Stockholm: Swedish Research Institute in Istanbul. Retsö, Jan. 1994. “ʾiʿrāb in the forebears of modern Arabic dialects.” Actes des premières journées internationales de dialectologie arabe de Paris. Colloque international tenu à Paris du 27 au 30 janvier 1992, ed. Dominique Caubet and Martine Vanhove, 333–342. Paris: INALCO. Retsö, Jan. 1995. “Pronominal state in colloquial Arabic. A diachronic attempt.” Dialectologia Arabica. A Collection of Articles in Honour of the Sixtieth Birthday of Professor Heikki Palva (Studia Orientalia 75), ed. Tapani Harviainen, 183–192. Helsinki: Finnish Oriental Society. Retsö, Jan. 1997. “State and plural marking in Semitic.” Built on Solid Rock. Studies in Honour of Professor Ebbe Egede Knudsenon the Occasion of His 65th Birthday April 11 1997, ed. Elie Wardini, 268–282. Oslo: Novus forlag. Retsö, Jan. 1997. “The Arab connection. Political implications of frankincense in early Greece.” Profumi d’Arabia. Atti del convegno, ed. Alessandra Avanzini, 473–480. Rome: L’Erma di Bretschneider. Retsö, Jan. 1999. “Nabatean origins – one again.” Proceedings of the Seminar for Arabian Studies 29: 115–118. Retsö, Jan. 2000. “Kaškaša, t-passives and the ancient dialects in Arabia.” Oriente Moderno 19 (80) n.s. (Studi di dialettologia araba, ed. Lidia Bettini): 111–118. Retsö, Jan. 2000. “Where and what was Arabia Felix?” Proceedings of the Seminar for Arabian Studies 30: 189–192. Retsö, Jan. 2002. “Das Arabische der vorislamischen Zeit bei klassischen und orientalischen Autoren.” Neue Beiträge zur Semitistik. Erstes Arbeitstreffen der Arbeitsgemein-

Bibliography of Jan Retsö

15

schaft Semitistik in der Deutschen Morgenländischen Gesellschaft vom 11. bis 13. September 2000 and der Friedrich-Schiller-Universität Jena (Jenaer Beiträge zum Vorderen Orient 5), ed. Norbert Nebes, 139–146. Wiesbaden: Harrassowitz. Retsö, Jan. 2002. “Legendernas drotting. Berättelsen om drottningen av Saba och dess ursprung.” Svensk Religionshistorisk Årsskrift 11: 135–159. Retsö, Jan. 2003. “When did Yemen become Arabia felix?” Proceedings of the Seminar for Arabian Studies 33: 229–235. Retsö, Jan. 2003. “Hur läser man Koranen innantill. En tolkning av elefantsuran.” Svensk Religionshistorisk Årsskrift 12: 103–111. Retsö, Jan. 2004. “Relative-clause marking in Arabic dialects: A preliminary survey.” Approaches to Arabic Dialects. A Collection of Articles Presented to Manfred Woidich on the Occasion of His Sixtieth Birthday (Studies in Semitic Languages and Linguistics 38), ed. Martine Haak, Rudolf de Jong, and Kees Versteegh, 263–273. Leiden: Brill. Retsö, Jan. 2004. “In the shade of Himyar and Sasan – The political history of preIslamic Arabia according to the ʾAyyām al-ʿarab-literature.” Arabia 2: 111–118. Retsö. Jan. 2005. “The number-gender-mood markers of the prefix conjugation in Arabic dialects. A Preliminary consideration.” Edzard and Retsö (eds.) 2005: 31–40. Retsö, Jan. 2005. “Arabs in pre-Islamic South Arabia.” Archäologische Berichte aus dem Jemen 10: 127-131. Retsö, Jan. 2005. “Arabia and the heritage of the Axial Age.” Axial Civilizations and World History (Jerusalem Studies in Religion and Culture 4), ed. Jóhann Árnason, Shmuel N. Eisenstadt, and Björn Wittrock, 337–358. Leiden: Brill. Retsö, Jan. 2006. “Arab.” Encyclopedia of Arabic Language and Linguistics, ed. Kees Versteegh, vol 1: 126–133. Leiden: Brill. Retsö, Jan. 2006. “Aramaic (Syriac) loanwords.” Encyclopedia of Arabic Language and Linguistics, ed. Kees Versteegh, vol. 1: 178–182. Leiden: Brill. Retsö, Jan. 2006. “Diathesis.” Encyclopedia of Arabic Laguage and Linguistics, ed. Kees Versteegh, vol. 1: 622–626. Leiden: Brill. Retsö, Jan. 2006. “Thoughts about the diversity of Arabic.” Edzard and Retsö (eds.) 2005: 23–33. Retsö, Jan. 2006. “The concept of ethnicity, nationality and the study of ancient history.” Topoi 14/1: 9-17. Retsö, Jan. 2007. “ʿAraba, Wādī.” Encyclopedia of Islam 3rd ed., vol. 1: 97. Leiden: Brill. Retsö, Jan. 2007. “Arabs (historical).” Encyclopedia of Islam 3rd ed., vol. 1: 73–78. Leiden: Brill. Retsö, Jan. 2007. “ʿAtīra.” Encyclopedia of Islam 3rd ed., vol. 1: 116–117. Leiden: Brill. Retsö, Jan. 2008. “Konstantinopel och den tidiga islamiska eskatologin.” Bysantinska sällskapet. Bulletin 26: 21–32. Retsö, Jan. 2008. “Wahb b. Munabbih, the Kitab al-tījān and the history of Yemen.” Arabia 3: 227–236.

16

Bibliography of Jan Retsö

Retsö, Jan. 2009. “Nominal attribution in Semitic: Typology and diachrony.” Watson and Retsö (eds.) 2009: 3–33. Retsö, Jan. 2009. “The m-suffix with the finite verb 3rd and 2nd person plural in Arabic dialects.” Zeitschrift für Arabische Linguistik 50: 39–50. Retsö, Jan. 2009. “Oxford wordpower.” Moderna Språk 103/1: 109–112. Retsö, Jan. 2010. “Status – eine vernachlässigte Kategorie der arabischen Grammatik.” Grammatik, Dichtung und Dialekte. Beiträge einer Tagung in Erlangen zu Ehren von Wolfdietrich Fischer, ed. Shabo Talay and Hartmut Bobzin, 65–71. Wiesbaden: Reichert. Retsö, Jan. 2010. “Arabs and Arabic in the age of the Prophet.” The Qurʾān in Context. Historical and Literary Investigations into the Qurʾānic Milieu, ed. Angelika Neuwirth, Nicolai Sinai, and Michael Marx, 281–292. Leiden: Brill. Retsö, Jan. 2011. “Classical Arabic.” The Semitic Languages. An International Handbook, ed. Stefan Weninger, 782–810. Berlin: de Gruyter Mouton. Retsö, Jan. 2011. “Petra and Qadesh.” Svensk Exegetisk Årsbok 76: 115–136. Retsö, Jan. 2011. “Constantinople and the early Islamic conquests.” Istanbul as seen from a distance. Centre and province in the Ottoman empire, ed. Elisabeth Özdalga, M. Sait Özervarli, and Feryal Tansug. Swedish Research Institute in Istanbul Transactions 20: 29–36. Retsö, Jan. 2012. “The Nabataeans – problems of defining ethnicity in the Ancient World.” Visions of Community in the Post-Roman World. The West, Byzantium and the Islamic World 300–1100, ed. Walter Pohl, Clemens Gantner, and Richard Payne, 73– 80. Farnham/Burlington, VT: Ashgate. Retsö, Jan. 2013. “What is Arabic?” The Oxford Handbook of Arabic Linguistics, ed. Jonathan Owens, 433–450. Oxford: Oxford University Press. Retsö, Jan. 2013. “Diathesis.” Encyclopedia of Hebrew Language and Linguistics, ed. Geoffrey Khan, vol. 1: 271–274. Leiden: Brill. Retsö, Jan. 2014. “The The b-imperfect.” Encyclopedia of Arabic Language and Linguistics Online, ed. Lutz Edzard and Rudolf de Jong. Leiden: Brill. Retsö, Jan. 2014. “The b-imperfect once again: typological and diachronic perspectives.” Proceedings of the Oslo–Austin Workshop in Semitic Linguistics. Oslo, May 23 and 24, 2013, ed. Lutz Edzard and John Huehnergard (Abhandlungen für die Kunde des Morgenlandes 88), 64–72. Wiesbaden: Harrassowitz. Retsö, Jan. 2014. “The contradictory revelation – a reading of Sura 27:16–44 and 34:15– 21.” Micro-Level Analyses of the Qurʾān, ed. Håkan Rydving (Acta Universitatis Upsaliensis religionum 34), 95–103. Uppsala: University Press. Retsö, Jan. 2015. “The problem of circumstantial clause combining (CCC) in Sabaean.” Clause Combining in Semitic. The Circumstantial Phrase and Beyond, ed. Bo Isaksson and Maria Persson (Abhandlungen für die Kunde des Morgenlandes 96), 297–364 Wiesbaden: Harrassowitz. Retsö, Jan. to appear. “Readings from the Qurʾān. Introduction and commentary to selected passages.”

Bibliography of Jan Retsö

17

3. Reviews Retsö, Jan. 1985. “Review of Norbert Nebes. 1982. Funktionsanalyse von kāna yafʿalu. Ein Beitrag zur Verbalsyntax des Althocharabischen mit besonderer Berücksichtigung der Tempus- und Aspektproblematik. Hildesheim: Olms.” Acta Orientalia 46: 175–181. Retsö, Jan. 1993. “Review of “Frank Henderson Stewart. 1988. Texts in Sinai Bedouin Law. With the assistance of Haim Blanc and Salāmih Ḥsēn. Part I: The texts in English Translation; Part II: The texts in Arabic. Wiesbaden: Harrassowitz.” Acta Orientalia 54 (1993) 179–182. Retsö, Jan. 1995. “Review of Tryggve M. Mettinger. 1995. No Graven Image? Israelite Aniconism in its Ancient Near Eastern Context. Stockholm: Almqvist & Wiksell.” Svensk Teologisk Kvartalsskrift 4: 178–180. Retsö, Jan. 1995. “Review of Aline Tauzin. 1993. Contes arabes de Mauritanie. Paris: Karthala.” Acta Orientalia 56: 230–233. Retsö, Jan. 1995. “Review of Manfred Ullmann. 1992. Das Motiv des Spiegels in der arabischen Literatur des Mittelalters. Göttingen: Akademie der Wissenschaften.” Acta Orientalia 56: 225–229. Retsö, Jan. 1998. “Review of Norbert Nebes. 1995. Die Konstruktionen mit /fa-/ im Altsüdarabischen. Syntaktische und epigraphische Untersuchungen. Wiesbaden: Harrassowitz.” Acta Orientalia 59: 271–277. Retsö, Jan. 1999. “Review of Hussein Abdul-Raof. 1998. Subject, Theme and Agent in Modern Standard Arabic: Richmond: Routledge/Curzon.” Acta Orientalia 60: 227– 233. Retsö, Jan. 2000. “Review of Kees Versteegh. 1997. The Arabic Language. Edinburgh: Edinburgh University Press.” Al-ʿArabiyya. Journal of the American Association of Teachers of Arabic 33: 123–130. Retsö, Jan. 2001. “Review of Rudolf de Jong. 2000. A Grammar of the Bedouin Dialects of the North Sinai Littoral. Bridging the Gap between the Eastern and Western Arab World. Leiden: Brill.” Acta Orientalia 62: 221–228. Retsö, Jan. 2004. “Review of Alexander Sima. 1999. Die lihyanischen Inschriften von al.ʿUdhayb (Saudi-Arabien) (Epigraphische Forschungen auf der arabischen Halbinsel Band 1). Rahnen /Westfalen: Verlag Marie Leidorf.” Zeitschrift für Arabische Linguistik 43: 90–92. Retsö, Jan. 2004. “Review of Helen Younansardaroud. 2001. Der neuostaramäische Dialekt von Särdä:rïd. Wiesbaden; Harrassowitz.” Orientalia Suecana 53: 197–201. Retsö, Jan. 2005. “Review of “Ignacio Ferrando Frutos. 2001. Introducción a la historia de la lengua árabe. Nuevas perspectivas. Zaragoza.” Collectanea Christiana Orientalia (CCO) 2: 474–481. Retsö. Jan. 2006. “Review of Eva Orthmann. 2002. Stamm und Macht. Die arabischen Stämme im 2. und 3. Jahrhundert der Higra. Wiesbaden: Harrassowitz.” Jerusalem Studies in Arabic and Islam 31: 343–355. Retsö, Jan. 2009. “Review of Hagith Sivan. 2008. Palestine in Late Antiquity. Oxford: Oxford University Press.” Journal of the Royal Asiatic Society, Series 3, 19: 1–4.

18

Bibliography of Jan Retsö

Retsö, Jan. 2011. “Review of Iwona Gajda. 2009. Le royaume de Himyar à l'èpoque monothéiste. L'histoire d’Arabie du Sud ancienne de la fin du IVme siècle jusqu'à l'avènement de l’Islam. Paris: AIBL & De Boccard.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 161: 476–480. Retsö, Jan. 2014. “Review of Jaeyoung Jeon. 2013. The Call of Moses and the Exodus Story: A Redactional-Critical Study in Exodus 3–4 and 5–13. FAT 2:60.” Tübingen: Mohr. Svensk Exegetisk Årsbok 79: 178–180. Retsö, Jan. to appear. ”Review of Peter Stein. 2012/2013. Lehrbuch der sabäischen Sprache 1. Teil: Grammatik; 2. Teil: Chrestomathie (Subsidia et Instrumenta Linguarum Orientis 4,1; 4,2). Wiesbaden: Harrassowitz.” Zeitschrift der Deutschen Morgenländischen Gesellschaft.

Canonical and non-canonical uses of the imperative in Slavic1 Silje Susanne Alvestad, University of Oslo2

1

Introduction

In the Slavic languages, the imperative – in a pre-theoretical sense of the word – can be used in a wide variety of contexts. In this paper I will demonstrate some of these uses based on the distinction between ‘canonical’ and ‘non-canonical’ imperatives familiar from the typological literature (see, e.g., Aikhenvald 2010). The paper is example-oriented and structured as follows. In section 2, I will narrow down the object of study and define what I take an imperative to be. I will also comment on the canonical-non-canonical distinction. In section 3, I will present some examples of canonical uses of imperatives, including speech act types associated with necessity and possibility, and imperatives in complex constructions. Subsequently, in section 4, I will show some examples from Slovene, a Slavic language that allows for embedded imperatives. Section 5 will be devoted to non-canonical uses of imperatives, one of which is familiar from Arabic too – namely, the so-called narrative use of imperatives (see, e.g., Palva 1977, 1984, and Henkin 1994). In section 6, I will briefly discuss two hypotheses regarding the origins of the non-canonical uses of imperatives in Slavic, and, finally, I will draw some conclusions in section 7.

                                                                                                                        1 It is well known that the honoré of this Festschrift is not only a Semiticist but also a Slavicist by training. This paper, which is based on my PhD trial lecture on The functional diversity of Slavic imperatives, is thus directed towards Jan Retsö the Slavicist. Puzzles still abound in Slavistics too, not least when it comes to imperatives. In this paper I will present and discuss a selection of examples from the following Slavic languages, for which the following abbreviations in parentheses will be used in the examples: Russian (Ru), Belarusian (By), Ukrainian (Uk), Polish (Pl), Bulgarian (Bg), Serbian (Sr), Croatian (Hr), Macedonian (Mk), Upper Sorbian (US), Slovak (Sk), Czech (Cz), and Slovene (Sn). 2

I am thankful to my informants – Irena Marijanović, Agnieszka Będkowska-Kopczyk, Volha Farmuhina, and Katja Brankačkec – for their always quick response and their helpful comments. Needless to say, all errors are mine.

20

2

Silje Susanne Alvestad

The object of study

In the theoretical literature, there are both “form-biased” and “function-biased” approaches to what imperatives are (see, e.g., Birjulin 1994, and Hamblin 1987, respectively). Taking a form-biased approach to imperatives in Slavic, for example, we would perhaps need to account for third person singular forms, such as the Russian pust’ ‘let’ and xot’ ‘even if’. We might also have to consider first person plural forms, such as the Russian pojdemte ‘let’s go’. Taking a functionbiased approach, on the other hand, we would perhaps need to take into consideration verb-less adverbial constructions, such as the Slovene Vsi ven! ‘Everybody out!’. We might also have to account for certain uses of modal verbs, as in the Slovak sentence Môžeš si vybrat’ ‘You can choose’. In this paper, however, I follow Kaufmann 2012 and take an imperative to be a form-function pair, a combination of a particular morphological form and a particular set of functions. More specifically, I will be concerned with second person singular or plural verb forms that are typically used to perform speech act types such as ordering, commanding, requesting, instructing, suggesting, permitting, inviting, etc. They have a directive use. Such addressee-oriented imperatives, imperatives that are addressed “to those who are to carry them out”, are typologically widespread, and for this reason they are frequently referred to as canonical imperatives (see, e.g., Aikhenvald 2010: 47, referring to Lyons 1977: 747). In the Slavic languages, though, second person imperative forms are found in other contexts as well. I already mentioned the narrative use. This use, and some other uses of the imperative verb form are typologically less frequent. However, while Aikhenvald (2010: 47, 396) uses ‘non-canonical imperatives’ to refer to commands directed at persons other than the addressee—that is, at first person (me or us) and third person (him, her, or they), I will use the term slightly differently. Specifically, I will use the term ‘non-canonical’ to refer to a set of non-directive, non-addressee-oriented uses of second person imperative forms. The already mentioned narrative imperative is one such second person imperative form. It occurs in non-directive use and its subject typically does not refer to the addressee, but to the first or third person. There is no obvious way of classifying the various non-canonical uses of the imperative in Slavic. A variety of approaches are taken in the literature. Typically, a small set of uses is considered for a small set of Slavic languages. A case in point is Jakab (2002, 2005), who takes into account what she refers to as contrastive and counterfactual conditional imperatives in Russian. Gronas (2006),

Canonical and non-canonical uses of the imperative in Slavic

21  

on the other hand, takes Bulgarian, Macedonian, and Bosnian, Croatian, and Serbian into consideration in addition to Russian, but only considers what he labels the historical imperative. Gronas’ (2006) historical imperative includes what is more frequently referred to as the narrative imperative, but also some other uses. All this is to say that there is no consensus in the literature as to how to categorize non-canonical uses of the imperative in Slavic. My objective here, though, is to give a little taste of the wide range of possible uses of the second person imperative form in Slavic languages. It should also be borne in mind that the list of non-canonical uses presented in this paper may not be exhaustive, in the sense that there might be more non-directive uses of second person imperatives in Slavic languages. Note, as well, that non-canonical imperatives are found in the Slavic languages to varying degrees. The narrative imperative does not exist in Polish and Sorbian, for example (cf. Gronas 2006: 92). Summing up, I will first provide some examples of canonical imperatives in Slavic – that is, second person imperative verb forms that are used to direct the addressee to perform a certain action. Before I present a selection of non-canonical imperatives – that is, second person imperative verb forms that have a nondirective use and where the subject is typically not the addressee, I will provide some typologically rare examples from Slovene. Specifically, Slovene is a Slavic language that allows for embedded imperatives. Given that embedded imperatives are a rare phenomenon cross-linguistically, they should perhaps be listed among the non-canonical uses. On the other hand, as we will see, they are, in a way, addressee-oriented and directive, which is why I decided to present them separately as neither canonical, nor non-canonical. 3 Canonical uses of the imperative in Slavic Canonical uses of the imperative include, first, speech act types associated with necessity, such as orders and commands. Such imperatives can typically be paraphrased by means of the performatively used modal must. The imperative construction Close the door immediately! is very similar in meaning to You must close the door immediately!, for example. Canonical uses of imperatives also include speech act types associated with possibility, such as permissions and invitations. Imperatives in this use can typically be paraphrased by means of the performatively used modals can or may. An imperative construction such as Have a cookie! seems very similar in meaning to You may have a cookie, for example.

22

Silje Susanne Alvestad

The third sub-group of canonical imperatives includes speech act types that are not easily linked to either necessity or possibility. Dares, for example, appear to have the opposite meaning of what they actually express. If A says to B, Come on, try and hit me!, B is unlikely to interpret the utterance as You must try and hit me!, or You may try and hit me!, for that matter. Similarly, imperatives such as Go on then. Do as you wish, which are sometimes referred to as concessive imperatives in the literature (see, e.g., Kaufmann 2012: 12), are, perhaps, rather expressions of the speaker’s resignation than her wish that the addressee perform the actions referred to. Still, such uses of the imperative are cross-linguistically frequent and can therefore be considered as canonical. The final sub-group of canonical imperatives consists of imperatives in complex constructions. Cases in point are imperatives in conditional conjunctions, such as Do your homework and you’ll pass the exam. It should perhaps be mentioned that there are not necessarily clear-cut distinctions between all the speech act types exemplified below. As pointed out by, inter alia, Eckardt (2011), there is rather a gradual continuum between, for example, commands and orders, orders and requests, requests and advice, etc. Although an important observation, it is not relevant for my purposes here. 3.1 Speech act types associated with necessity: COMMAND, ORDER, REQUEST (1) Cz: – Vyjděte všichni do dvora! (COMMAND) (ParaSol.3 Ostrovskij: Kak zakaljalas’ stal’ [How the Steel was tempered](henceforth KZS)) “Everybody go outside!” (2) US:

– Wzmiće sej hišće někoho a přehladajče komandanturu a zastaranstwo. (ORDER) (ParaSol. Ostrovskij: KZS) “Take someone along with you and see that the Commandant’s office and the rear services are in proper shape.”

(3) Pl:

– Nie rusz, bandyto, psie! (WARNING) (ParaSol. Ostrovskij: KZS) “Don’t touch it, you bandit, you dog!”

                                                                                                                        3 The ParaSol corpus is “a parallel aligned corpus of Slavic and (some) other European languages”, cf. http://www.slavist.de.

Canonical and non-canonical uses of the imperative in Slavic

(4) Mk:

23  

Потоа месото внимателно извадете го од маста.4 (INSTRUCTION) Then, take the meat carefully out of the fat.

(5) Sr/Hr: Ne ostavljaj nas, oče. Do not leave us, father. (Galton 1976: 237f.)

(PRAYER)

3.2 Speech act types associated with possibility: PERMISSION, INVITATION, ADVICE (6) US: Tuž dźě wostań doma, hač ći noha zažiła.

(ADVICE)

(Galton 1976: 237) So stay at home as long as your leg is not healed. (7) Ru

— Мы завтра собираемся сделать прогулку в горы, — отвечала Кити. — Что же, поезжайте, — отвечала княгиня, … (PERMISSION) (RuN.5 Tolstoj: Anna Karenina) “We're meaning to make an expedition to the mountains tomorrow,” answered Kitty, “Well, you can go,“ answered the princess, …

(8) By:

Дома буду я адна. Прыходзь, Паўліх, мы будзем чытаць вельмі цікавую кнігу Леаніда Андрэева – “Сашка Жыгулёў”. (INVITATION) (ParaSol. Ostrovskij: KZS) I’ll be home all alone. Come over, Pavlusha! I have a very interesting book we could read – Leonid Andreev’s “Saška Žigulev”.

3.3 Other speech act types (9) Cz:

Buď zdráv! (WELL-WISH) (ParaSol. Andrić: Na Drini ćuprija [The Bridge over the Drina]) Stay healthy!

(10) a. Ru Βыиграй миллион! Win a million!

(ADVERTISING)

                                                                                                                        4 The example is from a Macedonian cookery book: Čačeska, Zorka. 2002. Makedonska kujna [The Macedonian Kitchen]. Draganić – Arabeska. The English translation is mine. 5

‘RuN’ refers to the Russian-and-Norwegian corpus, cf. http://www.hf.uio.no/ ilos/ english/research/projects/run/corpus/.

24

Silje Susanne Alvestad

b. By Βыйграй мільён!

(ADVERTISING)

c. US: Dobudźće 2 milionaj dolarow! d. Pl: Wygraj 2 miliony dolarów! Win 2 million dollars! (11)

US: Prošu, njeměj dalšu vazu złamanu … (WISH ABOUT THE PAST) (A mother on her way home, thinking about her child:) Please, don’t have broken another vase…

(12) a. US: Prošu budź bohaty… b. Pl: Proszę, bądź bogaty…

(ABSENT WISH)

c. Ru: Будь богатым … (On one’s way to a blind date:) Please, be rich … (13)

Bg: – Хайде де, опитай се да ме удариш! (DARE) (ParaSol. Rowling: Harry Potter and the Philosopher’s Stone) “Go on then, try and hit me!”

(14)

Ru: – Вы грубиян! – закричал Толстый. Вас гнать надо! Я на вас докладную подам! – Ну и подавайте, мрачно сказал Корнеев. – Займитесь любимым делом. (А. и Б. Стругацкие) (Xrakovskij 1988: 286, his (68)) (CONCESSION) “You boor!” Tolstyj shouted. “I’ll have to drive you away! I’m going to report you!” “Go on. Report me!” Korneev said gloomily. “Do as you wish.” (A. and B. Strugackij)

3.3 Imperatives in conditional conjunction and disjunction In the theoretical literature on imperatives it is frequently pointed out that imperatives often occur in various types of conditional constructions. Kaufmann (2012: 219–252), for example, thus discusses imperatives in conditional conjunction and disjunction. She distinguishes between three main types. Constructions such as Do your homework and you’ll pass the exam, are referred to as Imperative-andDeclarative constructions, or IaDs, (Kaufmann 2012: 219, following von Fintel and Iatridou 2009). There are two types of IaDs, and the one mentioned is considered as Type I. An example of a Type II IaD is this: Take one step closer, and I’ll blow your head off. The imperative serves as the antecedent in a conditional

Canonical and non-canonical uses of the imperative in Slavic

25  

clause in the case of both Type I and Type II IaDs, but in the latter case the imperative is used to perform what Eckardt (2011: 210) refers to as an anti-directive act. In the Type II IaD above, the imperative clause, Take one step closer, tells the addressee what not to do. The idea here is that IaDs appear to convey approximately the same meaning as if…, then… conditional clauses. Do your homework and you’ll pass the exam seems to mean almost the same as If you do your homework, (then) you’ll pass the exam. Similarly, Take one step closer, and I’ll blow your head off appears to be almost synonymous to If you take one step closer, (then) I’ll blow your head off. In IaDs, imperatives are involved in conditional conjunctions. Imperatives may, however, occur in conditional disjunctions as well. Thus, constructions such as Get out of my house, or I’ll call the police are referred to as IoDs, Imperative-or-Declarative constructions in the literature (see, e.g., Kaufmann 2012: 219, following von Fintel and Iatridou 2009). Below I will provide examples from Slavic languages of Type I and Type II IaDs as well as IoDs, but also a fourth kind of complex construction—namely, a case in which the imperative is the consequent in a conditional clause. (15)

Sk: – Prečítajte si román Ovad, dozviete sa. (IaD Type I) (ParaSol. Ostrovskij: KZS) “Read the novel ‘The Gadfly’ and you’ll find out!”

(16)

Mk: Даj му прст, ќe побaрa цeлa рaкa. (IaD Type I) Give him an inch and he will take a mile. (Lit.: Give him a finger, (and he) will take the whole hand.) (Jakab 2005: 300, from Hacking 1997: 214)

(17)

Mk: Нe доjди, пa ќe видиш штo тe чeкa! (IaD Type II) Don’t come, and you’ll see what’s in store for you! (Galton 1976: 239)

(18) a. Ru: Подойди ближе и я тебя застрелю. (IaD Type II) I you shoot.PF.PRS come.IMP.2SG closer and Come closer and I’ll shoot you. (Jakab 2005: 324, her (40a)) b. By: Падыдзі бліжэй і я цябе застрэлю. c. US: Pój bliže a ja tebje zatřělu. d. Pl: Podejdź, a cię zastrzelę!

26

Silje Susanne Alvestad

(19)

Uk: – Злізай, не то міліцію покличу!

(IoD)

(ParaSol. Bulgakov: Master i Margarita [The Master and Margarita]) “Get down, or I’ll call the police!” (20) a. Uk: – А, - заникуючись, проговорив він, – а якщо вони знову теє ... – Гм ... – задумався артист, - ну, тоді приходьте до нас знову. (IMPERATIVE CONSEQUENT IN CONDITIONAL CLAUSE) (ParaSol. Bulgakov: Master i Margarita [The Master and Margarita]) “A - and...”, he said, stammering, “and if . . . [the banknotes] turn into [candy wrappers] again…” “Hm...,” the artiste pondered, “well, then come to us again.” b. Sk: – Aa . . . – zajachtal, – a ak sa zas . . . oné . . . – Hm . . . - – zamyslel sa umelec, – no, tak príďte zase k nám. 4

Embedded imperatives in Slovene

Embedded imperatives are a typologically rare phenomenon (see, e.g., Sadock and Zwicky 1985),6 but in one Slavic language they exist: Slovene. According to Rus (2004: 12), the verbs in Slovene that typically embed imperatives are reci ‘say’, vztrajati ‘insist’, ukazati ‘order’, svetovati ‘suggest, and opozoriti ‘warn’. Embedded imperative clauses in Slovene may occur as embedded argument clauses, as in (21), embedded restrictive relative clauses, as in (22), or as nominal complement clauses, as in (23) (cf. Rus 2004: 11). (21)

Sn: Mama pravi, da mother says that

jo her

dobro poslušaj! well listen.IMP.2SG

Mother says you should listen to her carefully. (Lit.: Mother says that listen to her carefully.) (Dvořák 2005: 11, his (23a))

                                                                                                                        6 Exceptionally, embedded imperatives can occur in Classical Arabic, e.g., fa-ʾašāra ʾilay-him ʾan iṯbutū ʿalā ṣalāti-him prayer-your.M.PL and-signal.PF.3M.SG to-them.M that be:firm.IMP.2M.PL at He [Muḥammad] gave them a sign to continue with their prayer. (Sīra Nabawīya, cf. Brünnow and Fischer [2008], Arabic p. 62, and W. Fischer 2006: § 414, Note 1.)

Canonical and non-canonical uses of the imperative in Slavic

(22)

Sn: To this

je avto, is car

ki ga prodaj which it sell.IMP.2SG

27  

čimprej. as soon as you can

This is a car that you must sell as soon as you can. (Lit.: This is a car that sell it as soon as you can.) (Rus 2004: 11, his (13c)) (23)

Sn: Zakaj te why you

moje my

opozorilo, da warning that

bodi previden, be.IMP.2SG careful

tako jezi? so make angry Why does my advice that you must be careful make you so angry? (Lit.: Why does my advice that you be careful make you so angry?) (Rus 2004: 11, his (13d)) 5

Non-canonical uses of the imperative in Slavic

I touched upon the classification problem with respect to the non-canonical uses of the imperative in Slavic in section 2. Here I will be a bit more specific. In the existing literature, both synchronic and diachronic approaches are taken and classifications based on syntactic as well as semantic and/or pragmatic criteria can be found. The accounts also vary when it comes to which Slavic languages they take into consideration and which types of non-canonical uses they include. Before I present the examples, I will give a brief overview of the literature I have consulted based on the four abovementioned parameters – that is, whether the approach is synchronic or diachronic, based on syntactic, semantic, or pragmatic criteria (or a combination thereof), what Slavic languages are accounted for, and what types of non-canonical uses are included in the analysis. The eight works I have examined are thus presented in the table below (Figure 1). (With respect to column five, note that different researchers may use one and the same label to refer to distinct sets of non-canonical imperatives. Similarly, distinct terms used by different authors may cover one and the same set of noncanonical imperatives.)

28

Figure 1

Silje Susanne Alvestad

Overview of the literature consulted on non-canonical uses of the imperative in Slavic

Work

Perspective

Criteria

Languages

Uses

BelyavskiFrank (1991) Comtet (1994)

Synchronic

Semantic and pragmatic Syntactic

Russian Serbo-Croatian7 Russian

Narrative

Daiber (2009)

Diachronic

Russian

Fortuin (2000, 2008)

Synchronic

Semantic and pragmatic Semantic

Gronas (2006)

Diachronic

Semantic and pragmatic

Israeli (2001)

Synchronic

Syntactic semantic

Russian Serbo-Croatian10 Bulgarian Macedonian11 Russian

Synchronic

and

Russian

Dramatic8 Optativic Concessive Conditional Hypothetical9 Narrative/Historical Necessitive Conditional Concessive Narrative Optativic Historical12

Pseudo-imperatives, including: Narrative Necessitive/Contrastive13 Conditional

                                                                                                                        7 ‘Serbo-Croatian’ corresponds to the modern standard varieties of Bosnian, Croatian, and Serbian. 8 Comtet’s (1994) dramatic imperative corresponds to what is more frequently termed the narrative imperative, although this term is also not clearly defined. 9 Comtet’s (1994) hypothetical imperative corresponds to, i.a., Jakab’s (2002, 2005) counterfactual conditional imperative. 10 See footnote 6. 11 Mentioned in the article are also Belarusian, Ukrainian, Czech, Slovak, Slovene, Polish, and Sorbian, but no examples from these languages are provided. 12 Most, but not all, of Gronas’ (2006) historical imperatives are cases of what is more often referred to as narrative imperatives. Gronas’ (2006) historical imperatives also, however, include cases that are elsewhere referred to as necessitive (in Trnavac 2006, and Fortuin 2000 and 2008, for example), counterfactual, or conditional imperatives, and briefly mentioned are also ”the imperative of concession” and ”the hypothetical imperative” (cf. Gronas 2006: 99). 13 Israeli’s (2001) pseudo-imperatives include examples of the type labelled necessitive by Trnavac (2006) and Fortuin (2000, 2008), which is one of the two types referred to as contrastive imperatives by Jakab (2002, 2005).

Canonical and non-canonical uses of the imperative in Slavic

Jakab 2005)

(2002,

Synchronic

Syntactic semantic

Trnavac (2006)

Synchronic

Semantic

and

Russian Macedonian Russian

29  

Conditional counterfactual Contrastive14 Counterfactual conditional Necessitive Conditional Concessive

My aim in this paper is to give a taste of the wide variety of non-canonical uses that the imperative in Slavic may have. Therefore I will not discuss here how these uses should best be categorized, or whether the labels used in the existing literature are appropriate or not. Instead I will allow myself to take a “maximalist” approach and save the theoretical issues for a later article. Now let us look at the examples. 5.1 Narrative imperatives Narrative imperatives are typically used to refer to sudden, unexpected actions in the past. The actions are unexpected from the speaker’s perspective (cf. Israeli 2001: 3) and with respect to situations described in a preceding clause (cf. Jakab 2002: 133). Narrative imperatives are one of two types referred to as contrastive imperatives by Jakab (2002, 2005). In most, if not all cases, the subject in narrative imperatives is either first or third person, singular or plural. According to Gronas (2006: 92), the narrative imperative is productive in Russian, Belarusian, Ukrainian, Serbian, Croatian, Bulgarian, and Macedonian, but rare in Czech, Slovak, and Slovene, and non-existent in Polish and Sorbian. (24) a. Ru: Вот я и высматриваю, чтобы знать, какой портретик повесить, чтобы не влипнуть в историю, а то, знаете, Герасим Леонтьевич, мой сосед, недосмотрел хорошо да возьми и вывеси Ленина, а к нему как наскочат трое: оказывается, из петлюровского отряда. (ParaSol. Ostrovskij: KZS) I’m keeping an eye open myself so’s to know what portrait to hang up. Wouldn’t care to get into trouble like Gerasim Leont’evič next                                                                                                                         14 Jakab (2005) splits the contrastive imperatives into two types. The first consists of cases that are more frequently referred to as narrative imperatives in the literature, while the second is composed of cases that Trnavac (2006) and Fortuin (2000, 2008) label necessitive imperatives.

30

Silje Susanne Alvestad

door. You see, he didn’t look out properly and had just gone and hung up a picture of Lenin when three men rushed in – Petljura men as it turned out. b. By: Герасім Леонцевіч, мой сусед не даглядзеў добра ды вазьмі і вывесі Леніна, а да яго я наскочаць ўтрох: (25)

Ru: Вдруг он подскочи и ударь меня. Suddenly he jump-up.IMP.2SG and hit.IMP.2SG me Suddenly he jumped up (lit.: jump up!) and hit (lit.: hit!) me. (Jakab 2005: 339, her (71), from Barnetová et al. 1979).

In Russian, narrative imperatives are often preceded by idomatic expressions such as voz’mi (da) (i), as in (24a), ”to emphasize the unexpectedness and abruptness of the action” (Jakab 2002: 133). Voz’mi is itself a second person singular imperative form derived from vzjat’ in the sense of vzdumat’ ‘to take it into one’s head’ (ibid.). 5.1.1 Narrative imperatives with iterative/habitual and/or intensive interpretation Belyavski-Frank (1991) identifies a special kind of narrative imperatives with iterative, or habitual interpretation in Serbian, Croatian, Macedonian, and Bulgarian. Gronas (2006) argues that this subtype of narrative imperatives can be found in Russian as well and observes that it often has an “intensive” character (op. cit.: 92). The example below is from Serbian. (26)

Sr: Pre rata nikako nisu letovali u gradu, nego kak naiđi to, oni ti lepo nabavi ogrev za zimu, uzmi decu, pa hajd na Zlatibor. Before the war there was no way they would spend the summer in town, instead as soon as summer came, they’d get wood together for the winter, … take the children, and go off to Zlatibor. (Belyavski-Frank 1991: 128, her (15), from Stevanović 1979)

5.2 Necessitive imperatives The non-canonical imperatives in this section are referred to as necessitive imperatives by Trnavac (2006) and Fortuin (2000, 2008) but considered to be just one of two types of contrastive imperatives by Jakab (2002, 2005). (The second type of contrastive imperatives are narrative imperatives, discussed in 5.1.) Necessi-

Canonical and non-canonical uses of the imperative in Slavic

31  

tive imperatives typically occur in complex constructions in which the preceding declarative clause describes a pleasant action, while the imperative clause refers to an undesirable action that is imposed upon the subject (cf. Israeli 2001: 5). The subject is typically the first person – ’I’, as in (28), or ‘we’, as in (27). Israeli (2001: 12) proposes the following semantic description of constructions involving necessitive imperatives: i) X is compelled to do something, even though ii) she doesn’t want to do it. Meanwhile, iii) Y, who belongs to the same set as X, does not have to do it, or iv) Y by her actions has caused X (to be compelled) to do it. Finally, v) the statement represents X’s point of view. (27) a. Ru: – Ты слыхал? Они наших вешают, а их провожай к своим без грубостей! (ParaSol. Ostrovskij: KZS) “Hear that?” he said. “They hang our people, but we have to escort (lit. escort!) them back to their own side and treat them nicely besides.” b. Pl: Oni naszych wieszają, a ich odprowadzaj do swoich bez brutalnego traktowania! c. US: Woni našich powěšeja, ale jich přewodźej k jich ludźom bjeze wšěch hrubosćow! (28)

Ru: Мои друзья работают, как проклятые, а я отдыхай My friends work like condemned but I rest.IMP.2SG у моря, наслаждайся природой. by sea, enjoy.IMP.2SG nature My friends work themselves to death and I have to rest (lit. rest!) by the sea and enjoy (lit. enjoy!) the nature. (Israeli 2001: 6, her (7))

In addition to the languages exemplified – Russian, Polish, and Sorbian – we find necessitive imperatives in Belarusian. According to my respondents, necessitive imperatives are unusual in Polish and productive in Sorbian. 5.3 Optativic imperatives

Optativic imperatives typically have a third person singular subject that refers to the deity, as in (29). Note that Bóh ‘God’ is not vocative in this case, but nominative.

32

(29)

Silje Susanne Alvestad

US: Pomhaj help.IMP.2SG

Bóh! God.NOM

(May) God help (us)! As evidenced by (30), though, optative-like imperatives may also have a second person subject. In (30), it is the second person plural. (30)

Ru: Провaлиcь

вы! you.PL get-lost.IMP.2PL If only you got lost!/Blast you!/Get lost! (Fortuin & Boogart 2009, 8, their (9))

5.4 Conditional imperatives Conditional imperatives occur in a type of complex constructions in which an imperative clause serves as the protasis and a declarative clause serves as the apodosis (cf. Israeli 2001: 21). Conditional imperatives have subjects in either the first or third person. This feature distinguishes them from IaDs and IoDs discussed in section 3.4. above. The Sorbian example in (31) is thus a typical manifestation of the conditional imperative. (31)

US: Ale But

jeno přińdź come.IMP.2SG only

tam here

chudy prošer nutř poor beggar.SUBJ.SG in

a požebri sebi: hnydom and beg-a-little.IMP.2SG REFL: immediately

so

čoła chmurja. REFL foreheads cover

But if a poor beggar comes in and begs a little, they will immediately make faces. (Fasske 1981, via respondent) (Lit.: Come a poor beggar and beg a little, and they will immediately make faces.) Israeli (2001) discusses conditional imperatives in Russian and states that the imperative clause in the protasis typically introduces a “disastrous event” that has “certain consequences for the participants in the apodosis” (op. cit.: 21). She has not found any contexts with positive implications and concludes that “[b]enign, likely or highly desirable contexts do not allow” for conditional imperatives and esli-conditionals must be used instead (op. cit.: 20). She illustrates this by contrasting examples such as a. and b. in (32) below.

Canonical and non-canonical uses of the imperative in Slavic

(32) a. Ru: Умри die.IMP.2SG

я

сегодня,

что

c

вaми

будет?

I

today,

what

with

you

becomes

33  

Were I to die today, what would happen to you? (Israeli 2001: 15, her (19a), from a letter written by A. S. Puškin to N. N. Puškina on June 28, 1834) b. Ru: *Умри я сегодня, моя жена опубликует my wife publish.PRF.PRS die.IMP.2SG I today, мои стихи. my poetry Were I to die today, my wife would publish my poetry. (Israeli 2001: 21, her (25a)) In (32b) the imperative protasis describes the same dramatic event as in (32a), but the situation referred to in the declarative apodosis is quite neutral. On Israeli’s account, this is why a conditional imperative is infelicitous. (33) below may be evidence to the contrary, however. The context in the protasis appears desirable, and the consequences for the participants in the apodosis appear to be posi-tive, yet we do find a conditional imperative in the protasis, and not a esliconditional. (33)

Ru: Договорись мы и все пойдет по-другому. everything go.PRF.PRS differently agree.IMP.2SG we and If we agree, everything will become different. (Trnavac 2006: 90, her (30))

We find conditional imperatives in Russian, Belarusian, and Sorbian, and perhaps other Slavic languages too. According to Jakab (2005: 299), conditional imperatives are impossible in Macedonian, however. 5.5 Counterfactual imperatives Counterfactual imperatives may occur with subjects of any person and number, but they only occur in the East-Slavic languages—that is, Russian, Ukrainian, and Belarusian (cf. Jakab 2005: 299-300).15 Counterfactual imperatives may be                                                                                                                         15 My respondents confirm the non-existence of counterfactual imperatives in Serbian, Croatian, Polish, and Sorbian.

34

Silje Susanne Alvestad

formed also from verbs that are infrequent in the canonical imperative due to their semantics (cf. Israeli 2001: 12). The latter is illustrated in (35) below. (34) a. Ru: Будь там be.IMP.2SG there

я, не I NEG

случилось бы happened MOD

этого. this

Had I been there, it would not have happened. (Jakab 2005: 318, her (28)) b. By: Будзь там я, не здарылася б гэтага. (35)

Ru: Принадлежи эта книга мне, this book me.DAT belong.IMP.2SG я был бы самым счастливым человеком на свете. I was MOD most happy man in world Had this book belonged to me, I would have been the happiest man in the world. (Israeli 2001: 12, her (14a))

5.6 Concessive imperatives – Type I First, note that concessive imperatives as discussed here are not identical to imperatives that are used to perform concession-like speech act types, as in (14) above. The imperative in (14) is directive, while the first, concessive imperative in (36) is not. Rather, the first imperative clause in (36) should be translated into English by an even if p-clause. It is not an order for the addressee to die. If it were, it would be nonsensical to issue orders to be complied with afterwards. The addressee cannot find an electrician and start the power plant after he is dead. Thus, the first imperative is a concessive imperative, meaning approximately even if you have to die doing it…, … Note, though, that we have a second person subject in this case. (36) a. Ru: Умри, а монтера найди и пусти электростанцию. (ParaSol. Ostrovskij: KZS) Even if you have to break your neck doing it: find an electrician and start the power plant. (Lit.: Die, but find an electrician and start the power plant.) b. By: Памры, а манцёраў знайдзі і пусці электрастанцыю. c. Uk: Умри, а монтера знайди й пусти електростанцію! d. Pl: Skonaj, a znajdź montera i musisz uruchomić elektrownię.

Canonical and non-canonical uses of the imperative in Slavic

35  

e. Sr Umri, ali montera nađi i stavi u pogon električnu centralu. f. Hr: Umri, ali montera nađi i stavi u pogon električnu centralu. g. US: Slakń mojedla, ale nańdź mi montera a čin, zo by ta milinarnja zaběhnyła! We find concessive imperatives of this kind in Russian, Belarusian, Ukrainian, Polish, Serbian, Croatian, and Sorbian, and perhaps more Slavic languages. 5.7 Concessive imperatives – Type II I have considered the kind of concessive imperatives exemplified below as a separate type because of their idiomatic character. Specifically, they are of the form, no matter where/what… + ni ‘NEG’ + IMP, and we find them in Russian and Belarusian. (37) a. Ru: В какую сторону ни глянь ― даль закрыта туманом. (Р. Б. Ахмедов. Промельки (2011), Бельские Просторы) (RNC.16 R.B. Axmedov. Promel’ki [Flashes] (2011), Bel’skie Prostory [The Bel’sk Area]) No matter which way you look, everything is covered in fog. b. By: У які бок ні глянь - далеч зачынена туманам. 6

The origins of the non-canonical (and canonical) uses of the imperative in Slavic: Hypotheses

There are two prominent hypotheses in the literature when it comes to the origins of the non-canonical (and canonical) imperatives in Slavic. I have placed ‘and canonical’ in parentheses here, since only one of the two hypotheses concerns the canonical uses as well. The first hypothesis can be referred to as the optative hypothesis and is espoused by, inter alia, Gronas (2006). The optative hypothesis holds that the Slavic imperative originates in the old Indo-European optative. More specifically, the narrative and the necessitive imperative originate in the Indo-European preterital optative, whereas the conditional, counterfactual, and concessive imperatives originate in the conditional optative.

                                                                                                                        16 ‘RNC’ refers to the Russian National Corpus, cf. http://ruscorpora.ru/en/.

36

Silje Susanne Alvestad

The second hypothesis can be referred to as the aorist hypothesis and is subscribed to by Daiber (2009), among others. According to the aorist hypothesis, the origin of the narrative and the necessitive imperative – but not the conditional, counterfactual and optativic imperative – is the aorist. Taking Russian as his point of departure, Daiber (2009) argues that these two non-canonical imperatives are one of the two relics of the former aorist in modern Russian. Specifically, “the most common forms of the Old Russian sigmatic aorist (second and third person singular) were homophonous with the imperative singular, and this phonological bridge was important in metaphorically identifying the imperative with the narrative use of the former aorist” (op. cit.: 23). The Russian imperative thus “took over some functions” from “the old aorist when the latter disappeared from Russian” (op. cit.: 25). In other words, the metaphorical use of the Russian imperative “starts precisely when the use of the aorist ends” (op. cit.: 27). If this were correct, however, as Gronas (2006: 90) points out, we would not expect to find narrative imperatives in languages where the aorist is still in active use, but we do– in Bulgarian and Macedonian, for example. Although this is an intriguing issue, I will unfortunately have to leave it for future research. 7

Conclusion: – Beware of the Declarative Fallacy!

My aim with this article was to give a little taste of the many different uses the imperative in Slavic may have. I would like to conclude with a quotation from Belnap (1990: 1) about the danger of focusing too narrowly on declarative clauses. He calls this the Declarative Fallacy. Thus, “(…) one should recognize that from the beginning there are not only declarative sentences, but, at least, both interrogatives and imperatives. The grammarians are right and those teachers of elementary logic that seem to have miseducated most of us are wrong: give all sentences equal time, and do not take declaratives as a paradigm for what can happen between full stops.”

References Aikhenvald, Alexandra Y. 2010. Imperatives and Commands. Oxford: Oxford University Press. Barnetová, Vilma et al., eds. 1979. Russkaja grammatika [Russian Grammar] I. Prague: Academia. Belnap, Nuel. 1990. “Declaratives are not enough.” Philosophical Studies 59 (1): 1–30.

Canonical and non-canonical uses of the imperative in Slavic

37  

Belyavski-Frank, Masha. 1991. “Narrative use of tense forms in Russian and SerboCroatian.” Slavic and East European Journal 35 (1): 115–132. Birjulin, Leonid A. 1994. Semantika i pragmatika russkogo imperativa [The Semantics and Pragmatics of the Russian Imperative] (Slavica Helsingiensia 13). Helsinki: Helsingin Yliopisto. Birjulin, Leonid A., and Viktor S. Xrakovskij. 1992. “Povelitel’nye predloženija: problemy teorii” [Imperative Sentences: Theoretical Issues]. Viktor S. Xrakovskij (ed.). Tipologija imperativnyx konstrukcij [Typology of Imperative Constructions], 5–50. St. Petersburg: Nauka.   Brünnow, Rudolf, and August Fischer. [2008] (8th commented edition by Lutz Edzard and Amund Bjørsnøs). Arabische Chrestomathie aus Prosaschriftstellern / Chrestomathy of Classical Arabic Prose Literature. Wiesbaden: Harrassowitz. Comtet, Roger. 1994. “L’impératif hypothétique en russe: Un cas de synonymie syntaxique.” Revue des études slaves 36 (3): 471-482. Available at http://www.persee.fr/ web/revues/home/prescript/article/slave_0080-2557_1994_num_66_3_6198 Daiber, Thomas. 2009. “Metaphorical use of the Russian imperative.” Russian Linguistics 33 (1): 11–35. Dvořák, Boštjan. 2005. “Slowenische Imperative und ihre Einbettung.” Available at http://web.fu-berlin.de/phin/phin33/p33t2.htm (The page numbers refer to this online version of the paper.) Eckardt, Regine. 2011. “Imperatives as future plans.” Ingo Reich et al. (eds.). Proceedings of Sinn und Bedeutung 15, 209–223. Saarbrücken: Universaar – Saarland University Press. Fasske, Helmut, and Siegfried Michalk. 1981. Grammatik der obersorbischen Schriftsprache der Gegenwart. Bautzen: Akademie der Wissenschaften der DDR. Fischer, Wolfdietrich. 2006. Grammatik des klassischen Arabisch. Wiesbaden: Harrassowitz. Forsyth, James. 1970. A Grammar of Aspect. Usage and Meaning in the Russian Verb. London: Cambridge University Press. Fortuin, Egbert. 2000. Polysemy or monosemy: interpretation of the imperative and dativeinfinitive construction in Russian. Doctoral Dissertation, Institute for Logic, Language and Computation, Amsterdam University. Amsterdam: ILLC Dissertation Series. Fortuin, Egbert. 2008. “Polisemija imperativa v russkom jazyke” [The Polysemy of the Imperative in Russian]. Voprosy jazykoznanija [Linguistic Issues] 1: 3–23. Fortuin, Egbert, and Ronny Boogaart. 2009. “Imperative as conditional: from constructional to compositional semantics.” Cognitive Linguistics 20 (4): 641-673. Available online at http://media.leidenuniv.nl/legacy/Imperative_as_conditional%20 Fortuin %20Boogaart.pdf (The page numbers refer to the online version of the paper.) Galton, Herbert. 1976. The Main Functions of the Slavic Verbal Aspect. Skopje: Macedonian Academy of Sciences and Arts.

38

Silje Susanne Alvestad

Gronas, Mikhail. 2006. “The origin of the Russian historical imperative.” Russian Linguistics 30 (1): 89–101. Hacking, Jane. 1997. “The Macedonian imperative: Reconciling exhortative and nonexhortative uses.” Balkanistica 10: 212–220. Hamblin, Charles L. 1987. Imperatives. Oxford: Basil Blackwell. Henkin, Roni. 1994. “On the narrative imperative in Negev Arabic and in Russian.” Journal of Semitic Studies 39 (2): 245–283. Huntley, Martin. 1984. “The semantics of English imperatives.” Linguistics and Philosophy 7 (2): 103–133. Xrakovskij, Viktor S. 1988. “Imperativnye formy NSV i SV v russkom jazyke i ix upotreblenie” [IPF and PF Imperative Forms in Russian and Their Usage]. Russian linguistics 3 (12): 269–292. Xrakovskij, Viktor S. 2001. Typology of Imperative Constructions. (LINCOM Studies in Theoretical Linguistics 9). Munich: LINCOM Europa. Israeli, Alina. 2001. “An imperative form in non-imperative constructions in Russian.” Glossos 1: 1-32. (Available online at http://slaviccenters.duke.edu/uploads/media _ items/israeli.original.pdf) Jakab, Edit. 2002. “Two cases of disagreement in Russian: contrastive imperatives and root infinitives.” Actes de l’ACL 2002, 132–144. Jakab, Edit. 2005. “Noncanonical uses of Russian imperatives.” Journal of Slavic Linguistics 13 (2): 299–357. Kaufmann, Magdalena. 2012. Interpreting Imperatives. (Studies in Linguistics and Philosophy 88). Berlin: Springer. Lyons, John. 1977. Semantics, vol. 2. Cambridge: Cambridge University Press. Mathiassen, Terje. 1996. Russisk grammatikk [Russian Grammar]. Oslo: Universitetsforlaget. Palva, Heikki. 1977. “The descriptive imperative of narrative style in spoken Arabic.” Folia Orientalia 18: 5–26. Palva, Heikki. 1984. “Further notes on the descriptive imperative of narrative style in spoken Arabic.” Studia Orientalia 55: 3-15, 379–391. Prokofieva, Rose. 1959. How the steel was tempered. 2nd ed. Moscow: Foreign Languages Publishing House. Translation of Nikolaj Ostrovskij’s Kak zakaljalas’ stal’, 1936. Rus, Dominik. 2004. “Embedded Imperatives in Slovenian.” Available at http://faculty. georgetown.edu/portnerp/nsfsite/Rus_Slovenian_Imperatives.pdf (accessed May 15, 2014). Sadock, Jerrold M., and Arnold M. Zwicky. 1985. “Speech act distinctions in syntax.” Timothy Shopen (ed.). Language Typology and Syntactic Description, vol. I, 155–196. Cambridge: Cambridge University Press. Stevanović, Mihailo. 1979. Savremeni srpskohrvatski jezik II: sintaksa [Modern SerboCroatian II: Syntax]. 3rd ed. Belgrade: Naučna knjiga.

Canonical and non-canonical uses of the imperative in Slavic

39  

Trnavac, Radoslava. 2006. Aspect and subjectivity in modal constructions. Ph.D. dissertation, University of Leiden, The Netherlands. von Fintel, Kai, and Sabine Iatridou. 2009. “Morphology, syntax, and semantics of modals.” Lecture notes for LSA Summer Institute.

New manuscript fragment of a Prolog, discovered in the University Library in Uppsala Antoaneta Granberg, University of Gothenburg

Abstract In 2014 a hitherto unknown Cyrillic manuscript fragment of a parchment leaf was discovered in the University Library in Uppsala Carolina Rediviva. The fragment has been used as cover for a printed book, Sententiæ et loci quidam insigniores, ex antiquioribus maximeque probatis poetis Ordine collecti, ac in libros quinq. digesti, Poznań 1583. The book belonged earlier to the library of the Jesuit Collegium in Riga and it was taken to Sweden as booty in 1622. The fragment consists of text in two columns, containing сло́во ѡ҆ ст҃ѣмъ Василїи Вели́цѣⷨ и҆ ѡ҆ Е!ѳрѣмѣ. По́вѣсть дш҃еполезна, a text for the 3rd of January in the long redaction of the Slavonic Prolog for the first half of the year (September – February). The fragment does not belong to any of the manuscripts preserved in the other manuscript fragments from Swedish libraries and archives. The fragment is reproduced here with short comments. The dating of the fragment, based on its paleographic, orthographic and linguistic features, proposed by the author of the article is the end of the fourteenth and (most probably) the beginning of the fifteenth century. The language is Russian Church Slavonic with some northwestern features that are usually observed in Pskovian and Novgorodian manuscripts.

1. Introduction1 Swedish libraries and archives have in their collections, in addition to a large group of valuable old Cyrillic manuscripts and early prints, a number of important Cyrillic parchment fragments that have already been highlighted in several studies.2 The number of known parchment fragments increased as a result of the project Digitalized Descriptions of Slavic Cyrillic Manuscripts and Early Printed Books at Swedish Libraries and Archives (2010 2013), which included 107 parch-

1

I am much obliged to Ralph Cleminson (Winchester) for his valuable suggestions and commentaries, and for his kindly help in improving this text.

2

One of the latest works on the topic, including an overview of the research field, is by Lars Steensland (Steensland 2013), whose long-standing investigation of the Cyrillic parchment fragments in Sweden has resulted in several important publications.

New manuscript fragment of a Prolog

41!

ment fragments from a total of 51 different parchment manuscripts.3 The number of Cyrillic parchment fragments in Sweden continues to rise and in 2014, a hitherto unknown Cyrillic parchment fragment was discovered in Uppsala’s Carolina Rediviva building, which houses the University Library.4 The fragment is used in a binding of a copy of Sententiæ et loci quidam insigniores (Sententiæ et loci 1583).5 This article aims to provide an identification and reproduction of the version of text preserved in the fragment, a presentation of its physical characteristics and concise comments on its orthographic, paleographic and linguistic characteristics. The article also seeks to highlight the provenance of the fragment. 2. Provenance The copy of Sententiæ et loci (Sententiæ et loci 1583) belonged to the library of the Jesuit Collegium in Riga. Jesuit life in Riga dates from 7 March 1583 and the library of the Jesuit Collegium probably dates from the same time. On 16 September 1621 the army of the Swedish king Gustaf II Adolf (1594–1632) conquered Riga and seized the collections of the library of the Jesuit Collegium (Walde 1916: 43–52).6 The year after, in 1622, the booty from the Jesuit Collegium in !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 3 This national Swedish project was accomplished by Alexander Pereswetoff-Morath (Stockholm University), Antoaneta Granberg (University of Gothenburg), Irina Lysén (Uppsala University) and Per Ambrosiani (Umeå University). The project was funded by the Swedish Foundation for Humanities and Social Sciences (Riksbankens Jubileumsfond). For information about the project and for a report on its results, see http://anslag.rj.se/en/fund/39532. Most of the investigation and the description of the parchment fragments in the project were done by Alexander Pereswetoff-Morath, see also his latest publication on this topic (PereswetoffMorath 2013). 4

Thanks to Håkan Hallberg, librarian at Carolina Rediviva, for drawing my attention to this fragment and for all his help with the investigation of the collection from the Jesuit Collegium in Riga.

5

Shelfmark Script. lat. coll. Poetæ 61:263.

6

Johannes Bothvidi (1575–1635) compiled an inventory of books from the Jesuit Collegium in Riga in 1622. The original of this inventory was not preserved but in the University Library in Uppsala there is a later copy from 1683. The rubric “Libri in Duodecimo”, p.25 in this copy, contains information about the copy of Sententiæ et loci quidam insigniores…, i.e. “Sententiæ & loci insigniores ex Poetis” (Bothvidi 1622: 25). In the course of investigating the provenance of this copy, the author of the

42

Antoaneta Granberg

Riga, consisting of around a thousand books, church utensils, liturgical objects and five Russian icons was moved to the University Library in Uppsala by order of the Swedish king.7 Small pieces of paper with a typed “R”, i.e. “Riga”, were placed by the Swedish bibliographer Isak Collijn (1875–1949) on all the books from the Jesuit Collegium in Riga kept in Carolina Rediviva. This is the second Cyrillic parchment fragment found in the collection of the Jesuit Collegium in Riga. The other fragment was also used in the binding of a printed book,8 and contains a text from the Acts and Epistles (Steensland 2005: no. 8). These two fragments do not belong to the same manuscript. 3. Condition and physical description The fragment consists of a single leaf, used for covering the binding of the book; it is upside down in relation to the book it covers. The other side of the leaf is not visible. The fragment is in relatively good condition, apart from the letters on the edge that have been erased. The skin above the raised bands of the spine of the printed book is worn and there are small holes on the edge between the spine and the two covers. The fragment still remains on the cover of the printed book.9 In modern times an endpaper has been glued inside the front cover and another one inside the back cover. All edges of the parchment leaf have been folded over 8–17 mm onto the inside of the covers. This part of the leaf is covered by endpapers and only very small parts of the letters are visible. A yellow paper label, 27×21 mm,

article went through the archive of Isak Collijn and discovered hitherto unknown notes about the collection of books taken as war booty from the Jesuit Collegium in Riga: “Jesuitkollegiets i Riga bibliotek. Bidrag till dess historia”. The results of the investigation of Collijns archive will be published by Antoaneta Granberg in a forthcoming article. 7 Gustav II Adolf founded Carolina Rediviva two years earlier, in 1620. For the history of the collections taken as war booty to Carolina Rediviva from different Jesuit libraries in the seventeenth century, and for different inventories of these collections, see Hagström Molin (2014: 309, 313–315); cf. also Arne (1915: 121–122) and Walde (1916: 46–49). 8 A copy of Peter Canisius’ Catechismus Graecolatinus, printed in Ingolstadt in 1614; shelfmark Obr.67:190. 9 See the photograph at the end of the article. I wish to express my gratitude to the staff of the University Library in Uppsala, Carolina Rediviva, for producing the photograph and for giving me a permission to publish it.

New manuscript fragment of a Prolog

43!

is glued on the upper part of the spine and it covers part of the text of the fragment.10 A small piece of paper with the letter R typed on it is placed on the lower part of the spine. Size of the covers: 156–157×100–103 mm; size of the spine 154×26–27 mm. The text preserved in the fragment consists of two columns. There are nineteen visible lines in the left column and twenty visible lines in the right column. The height of ten lines is 88–91 mm, line length of the preserved part of the text is 103 mm (left column) and 118–119 mm (right column). The height of the letters is 5 mm, and the size of the written area is 235×175 mm. Traces of text before the first and after the last of the extant lines indicate that the leaf in its original size contained more text. The text of the first column extends across the back cover from the turn-in at the fore-edge as far as the fold between the board and the spine, and the text of the second column from halfway across the spine as far as the fore-edge of the front cover. Thus (apart from the first and last extant lines of each column which are on the fold at the top and bottom, and the label mentioned above), the damage to the text is confined to the beginnings of the lines of both columns. 3.1 The Left Column (A) The upper, lower and left edges of the written area of the left column are on the turn-ins and, as mentioned above, are now covered by the modern endpaper. This has obscured the first characters of all nineteen lines preserved, except line no. 5. The characters on the crease are erased and only a few small parts of letters are visible inside the cover. Line no. 1 is almost invisible and the remains of only a few letters are legible. Line no. 19 was also damaged as a result of the folding but the remains can still be read. 3.2 The Right Column (B) In the right column, the text on line no.1 is now placed on the upper folded part of the leaf: three letters are fully visible inside the upper part of the spine and the rest of the letters from this line are only partly visible. Lines nos. 2 and 10 have been preserved without any damage. The number of characters per line is circa 20–24. The parchment in column (B), lines nos. 11 to 15, had been stitched before the text was written, the text being continuous. The paper label glued on

10 Many books from the Jesuit Collegium in Riga do have such label on the spine.

44

Antoaneta Granberg

the spine of the binding has covered the beginning of lines nos. 16 to 18 in the right column and this part of the text is not visible. . Paleographical features This is just a short comment on the paleographical features of the fragment. It has been written in a semiuncial (poluustav) that is typical for the end of the fourteenth and the first third of the fifteenth century: •

ж with high central stem;

•

cross-beams: o high inclined cross-beam for the jotized vowels ѥ and ю; high less inclined, almost horizontal cross-beam for ꙗ; o и is written more often with an inclined cross-beam in the middle and, on a few occasions, with a horizontal cross-beam; o н is written with а cross-beam that begins from the top of the first stem and ends in the middle of the second stem; м with sagging pretty sharp “belly”;

• •

ц, the descender continues the right-hand stroke without any interruption at the base-line; ꙁ with sharply curving lower part; the middle part of ѡ is short and the upper parts curve towards the middle; the two straight strokes in the upper part of ч join the stem in the middle of the letter; the letter is almost symmetrical and the stem is inclined;

• • •

ѣ with tall stem and a small loop; е with upwardly inclined beam in the middle; х with shorter upper part.

• • •

. Orthographical features

ѿ is used in the preposition отъ and in forms of the word отьць and the derivative отьчьскъ. ѡ is used in initial position: ѡбрѣтох⁞[о|мъ] A4+5, ѡбрѧщеть A17, ѡг⁞[ъ|нь] A15+ , ѡдиною A10 (cf. ѥдиною A18); ѡ is used in the middle of the word, after a prefix: въѡ|ружаѥт сѧ B3–4. There is no broad ѻ. The digraph оу is used in initial position: оуже A3, оумножи сѧ А5 and оучененьꙗ (sic!) В15; in the middle and at the end of the word, у have been used: въѡ|ружает сѧ B3–4, гумна A6, живущихъ A13, меду B17, мучень|[ꙗ] B16–17, противу 16

B8, труду B8, мьзду B8 and сущимъ A7.

New manuscript fragment of a Prolog

45!

Jotized ѥ and ꙗ are used at the beginning of the word and after vowels; non-jotized e and ѧ are used after consonants. •

The original ъ: weak ъ in word-final position maintained except when followed by a clitic (which combines with the preceding word to form a “prosodic word”): въ B4,.ихъ А8, хощемъ A3; but. в A12, B5, въѡ|ружа т сѧ B3–4, подъкопаѥт сѧ A15,

исто|[м]⁞(и)л A2–3; • •

ъ/o alternation: тъ B8, тo B6, B9, кровь B11; во А14, воскрньꙗ B5, во|сприꙗти B8–9; in several cases ъ has been spelled with ь: in the preposition вь B13 (followed by a front vowel in the next syllable: вь неиже);

The original ь: • weak ь in word-final position is orthographically preserved in most of the cases, e.g. заповѣдь B14, [хрⷭDт]⁞омь A18, тать A11,. 15, ѿц҃ь A8; this is also the reason for inserting line-final jer in приѥмь|лющи B11–12 and, most likely, in пь [роче] ѥ A11–12, too 11 •

ь in suffixes has been preserved on most of the occasions, e.g. жит(ь)|[ѥ] A5– 6, воскрⷭDньꙗ B5, причастьѥ B6, батьство B7, а пⷭDльскаꙗ B14, and in the comparative degree form [сл]ажьша B17;

The abbreviated words are written with: (1) Titlo “ G ” with a short vertical beam in the middle is used for abbreviated words without a superscript letter and for the numbers, e.g.

ѿць A8 and днии ·т· и ·ѯ ·и ·е A19. (2) Pokrytie “ ҇ ” is always used above the superscript letter in abbreviated words with a superscript letter, e.g. чтⷭDотѣ B5 and чтⷭDнѣѥ B10. “ ̋ ” is used above each omega in initial position (but not above ѿ) and once after a prefix, e.g. ѡ̋брѣтох⁞[о|мъ] A4+5, and въѡ̋|ружаѥт сѧ B3–4. According to the paleographical features, in combination with the orthographical features, the most probable date should be the first third of the fifteenth century.12 11 .

Steensland (2013: 169 and the examples under 2.д .

12 The observations in this article on the paleographical features in connection with their chronology are based mostly on the publications by Poppe (1976) and Turilov (1993, 2013). I am deeply grateful to Anatoly Arkadyevič Turilov at the Department of Medieval History at the Institute of Slavic Studies of the Russian Academy of Sciences, for his expertise on the paleographic questions concerning this fragment.

46

Antoaneta Granberg

6. The language of the fragment he language of the fragment exhibits common Russian Church Slavonic features but also some north-west Russian features. Among the expected Russian Church Slavonic features should be mentioned: • The absence of the nasal vowels ѫ and ѭ and their substitution by у and ю, e.g. сущимъ A7 and вѣрою B12. In the case of ѧ, it is replaced by [a], written by • •

• • •

•

ѧ, ꙗ or а: ѡ̋ брѧщеть А17, причастьѥ B6, ча|[да A8–9,.[р]⁞ѣⷲа A2, лѣ|нивꙑꙗ B2–3; ѡMдиною А10 (cf. ѥдиною А18); the omitted jers reflect their loss in weak position, in the middle of the word, e.g. кт(о) B2, много A9, настра|[да]⁞(в)ше A9–10.(cf. the orthographically preserved ъ in a prefix: подъкопаѥт сѧ A15); ь in the root maintained on some occasions, e.g. дьꙗволъ А12, B3, мьзду B8, пожьжеть А16, and it has been omitted on several other occasions, e.g. (д)ни A14, что A2, and also in all three forms of the pronoun вьсь used, e.g. всего B10;

ь in front of j does not develop into и, e.g. дьꙗволъ А12 and B3 *dj > ж and not жд: [сл]ажьша B17; the ending of verbs in the third person singular present (future simple) is, without any exceptions, -ть and not -тъ, e.g. пожьжеть A16, ѡ̋брѧщеть A17,. имѣѥть B2; the prefix пре- instead of прѣ-: пребꙑти B4.

There are few regional language features in the fragment: • c–č merger, cokan´e on four occasions: пррⷪDцьскаꙗ B15, цаса А14, ѿце А2, ѿцьска B14 •

рьку А18 with ь (and not e); this could be explained as a reduction, but such explanation is not unproblematic because of the position of ь – the reduction usually happens in the position after an accented syllable, e.g. иже → ижь.13 In manuscripts from Novgorod, o and ъ are often used interchangeably, as are e and ь. The e/ь alteration in this stem, known from other Slavic languages, migh have an actual phonological (and not purely orthographical) basis, too 14

13 About the absence of e in ркꙋще, рQкꙋще as a feature of manuscripts of Pskovian origin see Steensland 2006: 21, 86 and the bibliography given in his publication. 14 What is linguistic and what is orthographical and where one draws the line is an eternal question. The text of the fragment is too short and there is not sufficient material to draw firm conclusions from.

New manuscript fragment of a Prolog

•

47!

The form сахарѧ B17,.with ѧ instead of a, is probably due to analogy – there are many more words in -арь in Old Russian than in аръ.

Most parchment fragments preserved in Swedish libraries and archives originnate from north-west Russia, the region of Pskov and Novgorod. This fragment is not an exception. The set of linguistic features of the fragment suggests that it was written in north-west Russia.15 This agrees well with the fact that it served as a binding material for a book from Riga. As mentioned above, the paleographical and orthographical features suggest that the fragment might date from the end of the fourteenth or the first third of the fifteenth century. As А. Poppe (Poppe 1976) and A. Turilov (Turilov 2013) have already pointed out, Novgorodian manuscripts, especially those having a traditional content, do not contain features of the second south Slavic influence up until the end of the first third of the fifteenth century. Compared to the north-east Russian manuscripts, the second south Slavic influence in the Novgorodian manuscripts is approximately a quarter of a century late. The second south Slavic influence during the fourteenth and fifteenth centuries is usually used as terminus ante quem in the dating of the manuscripts.16 The fact that there are no features from the second south Slavic influence in the fragment is not an obstacle to dating it as late as the first third of the fifteenth century. 7. The Contents In spite of all the corruptions in the text, the contents can be identified with certainty. This fragment is a witness to a part of a text for the 3rd of January in the Prolog. The text is found only in the long redaction of the Prolog for the first half of the year (September – February). In the Great Menaion Reader (Velikija Minei Četii 1910: cols 138–141), the text is entitled (В тои жe дн҃ь) сло́во ѡ҆ ст҃ѣмъ Василїи

Вели́цѣⷨ и҆ ѡ҆ Е!ѳрѣмѣ. По́вѣсть дш҃еполезна.17 The Slavonic source of this text is not known, nor is the Byzantine source (Prokopenko 2011: 413).

15 It should be noted that the language of the fragment contains only a few regional features. Most of the features that are common to the manuscripts from Pskov and Novgorod, are absent from the fragment, e.g. the Novgorodian change *zdj and *zgj > жг, *vj > *vl’ > *l’, the preposition оу > оув in front of vowels: оув одра, etc. 16 For a comprehensive review of the status praesens of the second south Slavic Influence, see Turilov (2010: 181 sqq., 235 sqq). 17 This text is usually placed after “Сборъ 70 Х(с)въ оученикъ”, see Prokopenko (2011: 413).

48

Antoaneta Granberg

This text is not included in the Synaxarium. Nor has it been included in the short redaction of the Prolog.18 Larisa Prokopenko presents many important arguments for the opinion that the long redaction of the Prolog was compiled in the twelfth century and the short redaction of the Prolog was compiled later, during the first half of the thirteenth century (Prokopenko 2011: 208–210).19! The fragment has preserved part of сло́во ѡ҆ ст҃ѣмъ Василїи Вели́цѣⷨ и҆ ѡ҆ ЕSѳрѣмѣ, without the beginning and the end. The comparison with the text preserved in manuscript and printed sources shows that the text of the fragment is not continuous. Between the end of the text in column (A) and the beginning of the text in column (B) there are approximately twenty lines of text missing. That means the visible part of this fragment is only half of a leaf. There are several Prolog fragments in the Swedish repositories,20 but the comparison with those shows that the newly discovered fragment from Carolina Rediviva does not belong to any of the manuscripts known from the other fragments. None of the other known fragments from Sweden contains this text or any other of the texts for the 3rd of January. 8. Principles for Reproducing the Text of the Fragment (а), i.e. reading of partially visible letters. [a], i.e. erased letters in the fragment, that have been reconstructed or taken from the Great Menaion Reader. |, i.e. the end of the line. ⁞, i.e. fold. There is no critical edition of сло́во ѡ҆ ст҃ѣмъ Василїи Вели́цѣⷨ и҆ ѡ҆ Е!ѳрѣмѣ. По́вѣсть дш҃еполезна, nor has it been the purpose of this article to provide one. Furthermore, the author of the article did not have access to the older manu18 See the edition by Krys´ko et al. (2010: lvii, 568–573), cf. Prokopenko (2011: 413). 19 In the discussion of the chronology of the Slavonic translation of the Greek Synaxarium, however, it must be born in mind that the Greek Synaxarium was most probably compiled at the very beginning of the tenth century, shortly after the death of patriarch Antonius II in 901 AD (Vander Meiren 1984). It should also be noted that William Veder has recently discussed the markup of the Prolog and, according to him, the Greek Synaxarion was translated into Old Church Slavonic as early as the tenth century and the Russian Cyrillic manuscript tradition is based on an earlier Glagolitic manuscript tradition (Veder 2014). 20 Cf. Steensland (2005: nos 23–44), Korobenko (2006), Krys´ko (2010: lxxvii), and Prokopenko (2011: 7).

New manuscript fragment of a Prolog

49!

scripts preserving the long redaction of the Prolog. For the purpose of this article, the text variant preserved on the fragment has been reproduced below together with only few observations comparing it with the Great Menaion Reader (VMČ, Velikija Minei Četii 1910: cols 138–141), a manuscript from the digital library of Trinity Lavra of St. Sergius (Tr 63), and one of the editions of the early printed Prolog (Prolog 1661).

References A. Printed editions Arne, Ture J. 1915. “Några ryska helgonbilder i svenska samlingar”. Fornvännen 10: 117–148. Hagström Molin, Emma. 2014. “Biblioteksmaterialiseringar: Krigsbyten, samlingsordningar och rum i Uppsala universitetsbibliotek under 1600-talet.” In Återkopplingar edited by Marie Cronqvist, Patrik Lundell and Pelle Snickars, 309–327. Lund: Mediehistoria. Korobenko, Larisa. 2006. “Från Pskov till Uppsala – så här!” In Jako blagopesnivaja ptica : hyllningsskrift till Lars Steensland, edited by Per Ambrosiani, 71–87. Stockholm: Acta Universitatis Stockholmiensis. (= Stockholm Slavic Studies, 32). Krys´ko et al. 2010: Крысько, Вадим Б., Прокопенко, Лариса B. and Желязкова, Веселка, editors. Славяно-русский Пролог по древнейшим спискам: Синаксарь (житийная часть Пролога краткой редакции) за сентябрь-февраль: Том I: Текст и комментарии. Москва: Издательский центр “Азбуковник”. Pereswetoff-Morath, Alexander I. 2013. “Klinckowströms samling: Stafsunds slotts och Kungliga bibliotekets kyrkoslaviska pergamentsamlingar och deras tillkomst”. Slovo (Uppsala) 54: 114–132. Poppe 1976: Поппэ, Aнджей В. “К истории романских дверей Софии Новгородской.” In Средневековая Русь. Светлой памяти Николая Воронина, edited by Георгий К. Вагнер, 191–200. Москва: Наука. Prokopenko 2011: Прокопенко, Лариса В. Древний славянский рукописный Пролог: история создания, редакции, бытование в XII-XIV вв. (сентябрьское полугодие). [s.n.]: Lap Lambert Academic Publishing. Prolog 1661: Про́логъ ѿ мцⷭ!а септе́врїа, до мцⷭ!а ма́рта. Москва: Печатный двор. Sententiæ et loci 1583. Sententiæ et loci quidam insigniores, ex antiquioribus maximeque probatis poetis Ordine collecti, ac in libros quinq. digesti. In usum studiosæ iuventutis, in Collegio Posnaniensi Societatis Jesu. Addita sunt sub finem quædam Christianorum Poetarum, de rebus sacris Carmina: et Hymni quidam Ecclesiastici. Poznań: Officina Typographica Ioannis Wolrabi. About this edition see The Universal Short Title Catalogue (USTC), ID no. 242476, http://ustc.ac.uk/index.php/record/242476 Steensland, Lars. 2005. “Trash and Treasure: Russian Parchment Fragments in Swedish Archives.” In Medieval Book Fragments in Sweden: an International Seminar in

50

Antoaneta Granberg Stockholm 13–16 November 2003, edited by Jan Brunius, 210–225. Stockholm: Kungl. Vitterhets historie och antikvitets akademien.

__________

. 2006. Если бы господь в гроб не был положен. Об едной малоизвестной толковой азбуке. Stockholm: Stockholms universitet. (= Acta Universitatis Stockholmiensis, 31).

__________

. 2013. “Фрагмент средневековой рукописи Минеи в Упсале”. Slovo (Uppsala) 54: 164–180.

Turilov 1993: Турилов, Анатолий А. “Предисловие.” In Дополнения к “Предварительному списку славяно-русских рукописных книг XV в., хранящихся в СССР”, edited by Анатолий А. Турилов and Н. А. Охотина, 3–38. Москва: Институт славяноведения и балканистики РАН. __________

. 2010. Slavia Cyrillomethodiana: Источниковедение истории и культуры южных славян и Древней Руси. Межславянские культурные связи эпохи средневековья. Москва: Знак.

__________

. 2013. “‘Революции’ в истории славянской письменности и проблемы эффективности палеографического метода.” In Палеография, кодикология, дипломатика. Cовременный опыт исследования греческих, латинских и славянских рукописей и документов. Материалы Международной научной конференции в честь 75-летия доктора исторических наук, члена-корреспондента Афинской Академии Бориса Львовича Фонкича, edited by Ирина Г. Коновалова, Джамиля Н. Рамазанова, Елена В. Казбекова, 310–325. Москва: Институт всеобщей истории РАН.

Vander Meiren, W. 1984. “Précisions nouvelles sur la généalogie des synaxaires byzantins.” Analecta Bollandiana 102, 3–4: 297–301. Veder, William R. 2014. “Markup in the Prolog”. Polata Knigopisnaja 39. [forthcoming] Velikija Minei Četii 1910: Великія Минеи Четіи. Январь. Тетрадь I. Дни 1—6. 15:1. Москва: Изданіе Археографической Коммиссіи. Walde, Otto. 1916. Storhetstidens litterära krigsbyten. En kulturhistorisk-bibliografisk studie. I. Uppsala: Almqvist & Wiksell.

B. Manuscripts Bothvidi, Johannes. 1622. Inventarium ver Jesuitkollegiets i Riga bibliotek, uppgjort av hovpredikanten, sedermera biskopen i Linköping Johannes Bothvidi. Uppsala University Library Carolina Rediviva, (M.S. / U. 271). Tr 63. Prolog for September–January, Пролог на сентябрь–январь; without beginning, without end, 327ff. Electronic resource, Собрание "Прочие" библиотеки МДА: http://old.stsl.ru/manuscripts/173-iv/52; ff.291v–293r: В тоиⷤ д͡нь слово ѡ с͡тѣмъ

васиⷧ!и и ѡ велицѣмⷨ! ефрѣмѣ. Повѣсть д͡шеполезнаꙗ.

New manuscript fragment of a Prolog

51!

(Column A) [и придоша къ] (ѿцю и) [р]⁞ѣⷲа · тꙑ что тако ѿце · исто [м]⁞(и)л нꙑ ѥси · оуже не хощемъ 5

[стр]⁞адати · и дни не ѡ̋брѣтох⁞[o21 мъ] зане оумножи сѧ жит(ь) [ѥ н]⁞аше гумна и точила и с(к) [о]⁞(т)и и кони и сущимъ по на(с) [до с]⁞ꙑти · реⷱD же имъ ѿць ихъ · ча [да] ⁞ мнѣсте ли много настра

10

[да]⁞(в)ше · ѡ̋диною ли пожаръ в(o) [зьм]⁞eть или вода или тать ·22 пь [роче]⁞ѥ23 врагъ дьꙗволъ в лѣно [сти] ⁞ живущихъ · но ищѣте24 то [го] ⁞ (д)ни25 и цаса · во26 тъ27 бо днь ни

15

[та]⁞ть подъкопаѥт сѧ · ни ѡ̋г⁞[ъ] [нь]⁞ пожьжеть · ни вода потоп(и) [ть . ]⁞ аще ѡ̋брѧщеть нꙑ готов[ꙑ] [с хрⷭDт]⁞омь · но и ѥще ѥдиною рьку [вꙑ] ⁞ (въ)28 лⷮDѣ днии ·т ·и ·ѯ· и ·е

not in VMČ, Tr 63, Prolog 1661. Differences between Tr 63 and 21 и дни не ѡ̋ брѣтох VMČ are not given in the notes, if the text of the fragment is following VMČ. 22 Not in VMČ. 23 прочее VMČ, прочье Tr 63, та́коже Prolog 1661. 24 ищете VMČ, Tr 63. 25 дн҃е VMČ, Tr 63. 26 в VMČ, вo Tr 63. 27 тѡи VMČ, ты Tr 63. 28 еⷭD в VMČ, есть в Tr 63, Prolog 1661.

52

Antoaneta Granberg

(Column B)

5

10

15

20

здн(омъ но и страда)[ньѥ .](ѥж)[е] кт(о) ремество имѣѥть · на лѣ нивꙑꙗ бо дьꙗволъ · не29 въѡ̋ ружаѥт сѧ · пребꙑти же въ чтⷭDотѣ · и в годъ стго воскрⷭDньꙗ приꙗти стоѥ причастьѥ · то ѥсть батьство неистощимо ѥ · тъ30 противу труду мьзду во сприꙗти · то дарове нетлѣни ꙗ и всего мира чтⷭDнѣѥ · тѣло и кровь сна биꙗ · ѥже приѥмь лющи31 · и32 вси вѣрою спсають сѧ лѣⷮD ѥсть цркꙑ33 вь неиже а пⷭDльскаꙗ заповѣдь · ѿцьска ꙗ оучененьꙗ34 · и пррⷪDцьскаꙗ [реⷱDнь]ꙗ · и стрпⷭDчьскаꙗ мучень [ꙗ · сл]ажьша сахарѧ · и меду35 · [кь не](и)же приходѧще по всѧ ча [сы п](оминающе смⷬDть и плака) ти сѧ36

29 Not in VMČ, Tr 63, Prolog 1661. 30 то VMČ, Tr 63. 31 прїемлюще VMČ, приемлюⷮD Tr 63. 32 Scribal error for приѥмь|лющии; not in VMČ, Tr 63, Prolog 1661. 33 цр҃ькви VMČ, церк҃вы Tr 63. 34 ен repeated twice. 35 са́хара и̂ семида́ла VMČ, Tr 63, Prolog 1661 (and the other early printed editions of Prolog). 36 Not in VMČ.

New manuscript fragment of a Prolog

UUB, Script. lat. coll. Poetæ 61:263, the cover of the book. Photograph by Uppsala University Library.

53!

Living together with the Jews: A Palestinian Arabic text from Jaffa Werner Arnold, Universität Heidelberg

1

Introduction

The text published here was recorded in Jaffa during a fieldwork campaign in 2001–2003, which was initiated by the late Rafi Talmon of the University of Haifa.1 The field research was completed in a research group at the Institute for Advanced Studies at the Hebrew University of Jerusalem from September 2003 to February 2004. The honored jubilarian also belonged to our research group on the Palestinian dialects, and I offer him with pleasure my contribution on Palestinian Arabic, recalling the beautiful time we spent together in Jerusalem. The speaker of the text is Amīn Saqḥafi, called Abu Durġām, in 2003, a 70 years old Muslim who came as a child with his parents from the village of Ǧabalīye to Jaffa. Jaffa was once a small city2 that grew rapidly during the time of the British Mandate in the first half of the last century, when thousands of farmers from the surrounding villages settled down to work in the that time very important harbor of the city. In spite of the fact that Amīn Saqḥafi lived nearly all his life in Jaffa, he preserved some features in his dialect which are typical of the village dialect which he spoke as a child. The vowels u and i in simple closed word-final syllable are not lowered to o and e, which is always the case in the dialect of Jaffa. Old *ǧ is only sometimes pronounced as ž, but *q is almost always replaced by the glottal stop. However, in the sentences 12 and 17 the pronounciation g occurs in the word bagaṛ (alongside baʾar in sentence 13). Therefore, the original dialect of Ǧabaliye – as expected – belongs, together with the dialects of Ṣummēl and Isdūd, to the southern group of the village dialects of the Tel Aviv area in                                                                                                                 1

The campaign was sponsored by the German Israeli Foundation within the framework of a survey of the Arabic Dialects in Israel together with my colleagues from the Universities of Erlangen, Haifa, and Jerusalem.

2

Cf. Arnold 2006.

 

A Palestinian Arabic text from Jaffa

55

which old k is preserved and old q shifted to g in contrast to the northern group in which old q shifted to k and old k is affricated to č.3 The questions in the text were asked by the young bank employee ʿAli Tamīm (T) from Jaffa. He speaks the normal city dialect of Jaffa.4 Hebrew words are marked by h...h 2 Arabic Text (1) ʿādit iḥna bniḥki, ṣaddiʾni, mā kunna ninda šabāb, wlād yahūd, illa: yā wlād ʿammna! yaʿni wlād ʿammna, hbin ha-dudímh, ʿaššān mā ḥada ylāḥiẓ ʿat-tāni fī zaʿal hēna, illa maḥabbe w-wudde w-ṣadāqa. (2) ṣaddiʾni, yā amīr ʿAli,5 lamman ana biddi arūḥ addāyan maṣāri mā arūḥ ʿind ʿarab, illa arūʿ 6 ʿind yahūd, aʾullu: biddi maṣāri addāyan minkum. (3) yiǧi yʾulli: ʾaddēš biddak? walad ismu X… aḷḷāh yirḥamu, ismu Xáyim… hū l-isim ismu Xāyim Ildištin 7, Xáyim Ildištin. (4) yiddīni 8 b-il-alf w-b-il-alfēn w-bit-talāte, mā fī bēni w-bēnu wṛāʾ, mā niktibš iwṛāʾ. xōd ʾadd-ma biddak! (5) yaʿni šū hal-ʿīše li-kānat, iḥna w-iyyāhum. ḥilwe, mā fī ʿalēha ḥaki. (6) amma hādi, masʾalit iḍ-ḍarb il-iǧatna min aḷḷa, hādi bass min aḷḷa, farraʾatna binātna. (7) iǧu l-banadīʾ 9 il-ʿaṭlīn, iš-šabāb il-ʿaṭlīn w-il-iwlād it-talfanīn, iǧu xarrabu ʿalēna il-mužtamaʿ. (8) lēš bitrūḥ hēn? lēš ibtīǧi hēn? fa farraʾūna, waḷḷāhi iḥna yaʿni hani nišváʾ lexá!h  10 innu samin w-ʿasal kunna maʿ wlād il-yahūd. (9) mā nʾuliš wlād yahūd illa wlād ʿammna, w-binġīb, w-bniḥki, afrāḥna afrāḥ wlād yahūd, ʿazāna yaʿni l-hxatsh… yaʿni bʿīd ʿanna, iza kān ṣār ʿindhum ʿaza, nitʿāza maḥḥum 11, yaʿni nōxud b-il-xawāṭir.                                                                                                                 3 Cf. Arnold 2013. 4 Cf. Arnold 2004. 5

Addressed is the bank employee ʿAli Tamīm.

6 = arūḥ. 7 Probably ‘Edelstein’. 8 adda/yiddi ‘to give’ is a rare word in the dialects of Palestine. It is not used in the dialect of Jaffa and must have its origin in Ǧabalīye. 9 Sg. bandūʾ, cf. now Hopkins 2013. 10 = ani nišbáʿ lexá ‘I swear to you!’

56

Werner Arnold

(10) w-hādi kānat ʿīše ḥilwe. w-il-yōm ṣaʿib il-wāḥad yʿabbir ʿan it-tāni, yiǧi yʾullak innu ana… biddu yiǧi yʾullak biddu yitwaǧhalli 12 l-yōm. (11) ēš-mā biddak tiḥki, biddu yʾullak, yaʿni biʿabbir tawaǧhun aw takāzub. fhimt? (12) yaʿni mubālaġa fī, amma lākin iḥna ṣaddiʾni, innu iḥna kunna mabṣuṭīn, iḥna wǧirānna. Bēt Yām, min Xolon, bi-Tall Abīb wāḥad ismu Irlix 13, tāǧir baʾar ikbīr, wHoladanki, tāǧir baʾaṛ ikbīr. (13) kunna rrūḥ 14 nōxud b-il-ʿašar ṛūs, b-il-xamastaʿšar ṛās bagaṛ, bgūr w-uġnūm. yaʿni l-ḥamdu li-llāh. (14) T: ēš kunt itsawwi inta ʾabl it-tamānye w-arbʿīn? (15) tāǧir! (16) T: kunt tāžer ġanam w-baʾar? (17) bagaṛ, mawāši, ā mawāši, kunut tāǧir bagaṛ w-mawāši yaʿni. w-la-mīn itbīʿ? kullayítu 15 ʿala wlād ʿammna. wlād ʿammna li-humme il-yahūd, w-il-ʿarab, inʾūl laḅaʿḍ: iwlād ʿammna, ʿašān mā ḥada yizʿal. (18) hāda bʾūl ʿanni: yahūdi, hāda bʾūl ʿanni: ʿarabi. biʾūlu: wlād ʿammna. (19) ʾaddēš yā ibin ʿammi biddak b-il… fi-hal-ʿiǧil aw fi-hal-šalīyit ġanam, tibʾa xamastaʿšar ṛās ġanam. mā nʾuliš xarūf wāḥad willa xarūfēn, nʾūl iš-šalye hādi ʾaddēš? tibʾa ʿašara xurfān, xamastaʿšar xarūf. (20) yiʾullu: itwakkal ʿala aḷḷa, illi bitǧību minnak bīr barakāt w-arsil. (21) kunna nǧīb il-xurfān min il-qura, iṭ-Ṭīre, Imm il-Faḥim, Galansawe, il-Lidd, irRamle, yaʿni min il-qura l-ʿarabīye. (22) Bīr is-Sabiʿ kān kull xamīs fī sūʾ. la-Bīr is-Sabiʿ, min yaʿni Israʾīl dāxil Israʾīl, i nžīb l-ibḍāʿa min hēn, w-inbīḥḥa. 16 (23) ʿād b-ʿīd il-bēsax nistanna min il-ʿīd la-l-ʿīd la-wlād ʿammna, irrūḥ inžīb ṣafqit 17 ġanam, w-fi-kull bēt min il-yahūd yidbaḥu hkēbesh, yaʿni xarūf, ʿašān il-bēsax, wnistanna min il… nifraḥ aktar min il-yahūd la-ʿīd il-bēsax.                                                                                                                                                                                                                                                                                                                                     11 = maʿhum. 12 twaǧhan/yitwaǧhan ‘to flatter’. 13 Probably ‘Ehrlich’. 14

= nrūḥ.

15 ʿAli Tamīm from Jaffa says kullayāto. 16 = nbīʿha. 17 šalye is a small flock, up to 30 animals, ṣafqa is a big flock.

 

A Palestinian Arabic text from Jaffa

57

(24) T: mnīḥ bass hāda btiḥki šū ismo il-… lamma btiḥki iḥki maʿā ʿarabi laʾannu ma baʿrafš iza biʿraf ʿibrāni hū, bižūz biʿrafš ʿibrāni hūwe. (25) mā bihimmiš ʿašān – imxarbiṭ, imxarbiṭha šwayye ʿarabi w-ʿibrāni – ʿašān yōxud ʿala l-lahǧe l-ʿibranīye, laʾinnu wlād ʿammna, bārak aḷḷāh fīhum, iḥna, lā samaḥ aḷḷāh, ʿumurna mā tʿaddēna ʿāla ḥadd. (26) ʾaʿadna bi-biyarātna, w-awwal mā daxalu Israʾīl ʿalēna šāyif, salām w-ʾamān. (27) mā maddu īdhum ʿalēna wala ḥakūnna: ēš fī ʿindkum? hay il-biyāra, ahlan wasahlan, ḍallkum ʾaʿdīn! (28) ysaǧǧlūna: yaʿni asamīkum! w-il-ḥamdu li-ḷḷāh rabb il-ʿālamīn. niḥmid aḷḷa ʿala hal-ḥāl w-kull ḥāl, w-inšā aḷḷā tirǧaʿ il-ʿāde il-ʾadīme. (29) w-bniṭḷuḅ min aḷḷāh yfukk is-salām mā bēnna w-bēn iwlād ʿammna. hāda hū. ʿādit ēš biddna niḥkīlak. 3

Translation

(1) Well we tell (you), believe me, we always called the young people, the boys of the Jews: Oh cousins! Oh cousins means (in Hebrew) bin ha-dudím, so that nobody would look at the other as if there is anger here, (but there is) only love and affection and confidence. (2) Believe me, honourable ʿAli, when I want to go and borrow money, I don’t go to Arabs, I only go to Jews, I say to him: I want to borrow money from you. (3) He comes and says to me: How much do you want? A guy called X…, the Lord have mercy upon him, he was called Xáyim… his name was Xáyim Ildištin, Xáyim Ildištin. (4) He gives me thousand, and two thousand and three (thousand) without a (written) contract between us, we didn’t write a contract. (He said:) Take as much as you like! (5) What a (good) life we had together with them. It was beautiful, there is no doubt. (6) But this problem of fighting, that has come to us from God, this is from God, it has separated us from each other. (7) The useless bastards came, the useless youngsters and corrupted boys, they came and destroyed coexistence for us.

58

Werner Arnold

(8) Why are you going here? Why are you coming here? And so they separated us, by God, we were, I swear to you, we were like butter and honey with the Jewish boys. (9) We don’t say ‘Jewish boys’, only ‘our cousins’. And we go out (together), and we speak (with each other), our weddings were (also) the weddings of the Jews, our bereavements, that is half… it means, Heaven forefend, if they have a bereavement we are in mourning with them, it means we offer our condolences. (10) And that was a beautiful life. But today it is difficult for somebody to express himself to the other, he comes and he says to you: I… he wants to come and say to you, he nowadays wants to flatter me. (11) Whatever you want to say, he will tell you, that is he’ll express flattering or lying. Do you understand? (12) That is there is exaggeration, but we, believe me, we were happy, we and our (Jewish) neighbours (in) Bat Yam, from Holon, in Tel-Aviv was somebody called Irlix, an important cattle dealer, and Holadanki, (another) important cattle dealer. (13) We used to go and to take 10 head of cattle, 15 head of cattle, cattle and sheep. Thank God! (14) T: What used you to do before (19)48? (15) (I was a) dealer! (16) T: You were a dealer in sheep and cattle? (17) Cattle, livestock, yes livestock, I was a dealer in cattle and livestock. And to whom to you sell? Everything to our cousins, our cousins who are the Jews, (they) and the Arabs, we call each other cousins, so that nobody will get angry. (18) Does one call me a Jew, or one call me an Arab? (no!) they say: cousins! (19) How much do you want, my cousin, for the… for this calf or for this flock of sheep consisting of 15 head of sheep? We didn’t say (we want) one sheep or two sheep, we said how much (does) this flock cost?, consisting of 10 sheep, 15 sheep. (20) He says to him: Put your trust in God, whatever you give me, from you rises a well of blessings. (21) We used to bring the sheep from the villages: iṭ-Ṭīre, Imm il-Faḥim, Galansawe, il-Lidd, ir-Ramle, that is from the Arab villages.

 

A Palestinian Arabic text from Jaffa

59

(22) In Beer Sheva there was a market every Thursday. To Beer Sheva, that is from Israel, inside Israel, we took the stock from here (to Beer Sheva) and sold it. (23) Well at Passover we wait from feast to feast for our cousins, we go and bring a big flock of sheep, and in every house of the Jews they slaughter a keves, that is a sheep, for Passover, and we wait from the… we look forward to the Feast of Passover more than the Jews. (24) T: O.K. but this one speaks, how do you say… when you speak, speak Arabic with him, because I don’t know whether he knows Hebrew, it could be, that he doesn’t know Hebrew. (25) It is not important, so that — I mix (the languages up), I mix up Arabic and Hebrew a little bit — so that he may pick up the Hebrew dialect (as well), because our cousins, God bless them, all our lives God forbid! we didn’t wrong anybody. (26) We lived in our orange grove, and as soon as Israel entered to control us, (there was) peace and safety. (27) They didn’t raise their hands against us nor did they say: What do you own? This is (your) orange grove, you are welcome, stay (where) your are living. (28) They registered us: that is (they said): your names! And thank God, the Lord of the worlds, we thank God for this (good) situation and every situation, and God willing the old custom will return. (29) And we beg from God, that he will redeem peace between us and our cousins. That’s all. Now what shall we tell you?

References Arnold, Werner. 2004. “Die arabischen Dialekte von Jaffa und Umgebung.“ In: Martine Haak, Rudolf de Jong, and Kees Versteegh (eds.). Approaches to Arabic Dialects. A Collection of Articles Presented to Manfred Woidich on the Occasion of his Sixtieth Birthday, 33–46. Leiden/Boston: Brill. Arnold, Werner. 2006. “Arabic Village and City Dialects in the Tel Aviv Area.“ In: Lutz Edzard and Jan Retsö (eds.). Current Issues in the Analysis of Semitic Grammar and Lexicon II. Oslo-Göteborg Cooperation 4th-5th November 2005, 34–39. Wiesbaden: Harrassowitz.

60

Werner Arnold

Arnold, Werner. 2013. “The Arabic Dialect of Isdūd.“ In: Ali Ahmad Hussein (ed.). Branches of the Goodly Tree: Studies in Honor of George Kanazi, 268–272. Wiesbaden: Harrassowitz. Hopkins, Simon. 2013. “On the Etymology of Arabic bandūq "bastard.“ In: Renaud Kuty, Ulrich Seeger, and Shabo Talay (eds.). Nicht nur mit Engelszungen: Beiträge zur semitischen Dialektologie. Festschrift für Werner Arnold zum 60. Geburtstag, 145– 150. Wiesbaden: Harrassowitz.

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sinai Rudolf de Jong, Netherlands-Flemish Institute in Cairo

As a tribute to the great Jan Retsö, this contribution presents texts in transcription and translation recorded from speakers of two different Bedouin tribes in Sinai: the Tayāha, who have their dīrah on the central Tīh-plateau in Sinai, and the Awlād Saʿīd, who are found in Wādiy Ṣlāf in the west of the southern tip of the Sinai Desert (see map below). After texts of the Taṛābīn of Nuweiba and of the ʿLēgāt of southwestern Sinai (see map below), which were published in De Jong 2013, the texts in this contribution are another illustration of the type of differences found between dialects in Sinai. The dialect of the Tayāha can be characterized as a typical group I dialect (see De Jong 2000, 2011); it is much like the dialect type of the Ḏ̣uḷḷām of the Negev Desert, decribed in Haim Blanc 1970 and that of the “Bd Bedouin” in Henkin 2010. The dialect of the Awlād Saʿīd has been classified as one of the dialects of group VII, spoken in southern Sinai (De Jong 2011). Together with most other dialects in Sinai, these two dialects have been concluded to form part of Palva’s (1991 and 2008) larger group of Northwest Arabian (Bedouin) dialects. The major accumulations of isoglosses found in southern Sinai coincide with the escarpment of the Tīh Plateau. These bundles of isoglosses1 distinguish                                                                                                                         1 Since isoglosses have been drawn along the borders of the dīrahs of the tribes, between every set of two tribes, they overlap exactly to form bundles. This produces a picture that would seem somewhat unrealistic when compared to dialect maps produced for areas in, for instance, rural Europe. In the situation in Sinai, however, every tribal collective forms a speech community in itself in which intertribal differences are few. When there are differences, these are generally found between different subdivisions (e.g. clans) of tribes, and are thus often also geographically identifiable, since members of one subdivision (like a clan) generally live in the same location. Having isogloss bundles coinciding exactly with a tribal border is therefore not only a practical solution to the problem of geographically represent-

62

Rudolf de Jong

the dialects of group I in the centre of Sinai from the dialects of mountainous southern tip of the peninsula (between group I to the north and groups VI, VII and VIII to its south (see De Jong 2011: maps 0 and 88 in the appendix). The texts in this contribution were recorded from speakers of the Awlād Saʿīd and the Tayāha. Bailey 1985 estimates the arrival in Sinai of the Awlād Saʿīd to have been in the 14th century C.E. In De Jong 2011: 7: “Aṭ-Ṭayyib 1993: 681–682 (see also 1997: 360–367) relates a story describing how the Awlād Saʿīd joined the tribe of Ṣawālḥah during their days in the Ḥiǧāz, after which they came to Sinai together. In ibid. it is also reported that a branch (named Awlād Sayf) of the Awlād Saʿīd are originally Masāʿīd.” The arrival in Sinai of the Tayāha is in Bailey 1985 estimated to have been between the 10th (perhaps earlier) and 13th centuries. In De Jong 2011: 6: “The Tayāha are a relatively large tribe. Aṭ-Ṭayyib 1993: 566 reports that they came to Sinai with the Banū Hilāl (of ʿAdnānī origin) and that they were among the first tribes to ‘settle’ on the Tīh plateau. After the Taṛābīn had arrived there, several wars were fought over control of the land. Sawārkah, Biliy, Rmēlāt, Samāʿnah are mentioned as allies of the Tayāha in these wars. For some time they were also allied with the Ḥwēṭāt against the Sawārkah. For further details on their history, presence in other countries etc., see ibid.: 565–570 and also aṭ-Ṭayyib 1997: 227–233.”

                                                                                                                                                                                                                                                                                                                                                            ing differences that are basically of a sociolinguistic nature, but such a representation – up to a certain degree – also reflects the geographical reality.

 

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sina

63

Map in De Jong 2011:372 (free after Bailey 1991 and 2009)

64

Rudolf de Jong

The speakers Speakers of the Awlād Saʿīd are: ʿŌdah (ʿŌ), 35 years old (school training: 2 years ʾIʿdādiy in Ṭūr, the capital of South Sinai governorate), Niṣṣāṛ (N) and Maḥmūd (M) (both appr. 60 years old) and Šēx Niṣṣār (Š), appr. 65 years old. All speakers are originally from Wādiy Ṣḷāf, level of school training of the latter three is unknown. Speakers of the Tayāha are: Mḥammad (appr. 35 years old); Aṃṃ Xiḏ̣r (appr. 65 years old). Mḥammad could read and write. Aṃṃ Xiḏ̣r had no school training. During these recordings my good friend ʿĪd al-Aṭraš (ʿĪ) was present. ʿĪd speaks the dialect of the Taṛābīn of Ṛās Ṣadr (abbreviated as TAṢ in De Jong 2011: chapter 3) and did most of the interviewing in these recordings. A note on transcription In the texts below anaptyctic vowels are indicated as ә (irrespective of their phonetic value2). Hesitations, like “ehm”, are indicated as ә:. The conjunction w and prepositions l and b are written separately (i.e. they are not connected to the following word). So-called “bukaṛa”-vowels3 are indicated in superscript. In the Saʿīdiy text, a superscript u is used to indicate strong velarization accompanying the 2nd p. sg. masc. pronominal suffix, as in ṛafagāuḳ “your travelling companions”. When the u is in its role as anaptyctic vowel (to break up clusters in a string v + CḳC + v), it is written on the line (i.e. not superscript) ( > v + CuḳC + v). Prosodic lengthening of vowels is indicated by a colon following the vowel (v + :). Velarization which cannot be directly (or easily) attributed to primary or secondary consonantal emphatics in the same word is indicated by a dot under the vowel (e.g. ạ̄ for IPA [ɑ:]), or under the consonant, as in e.g. -ḳuw. Unclear passages are indicated between square brackets [ and ]. Speech pause is indicated as …

                                                                                                                        2 For a description of the phonetic quality of anaptyctic vowels in this dialect, see De Jong 2000:128-131. The conjunction әw (i.e. when w is preceded by the anaptyctic vowel) will often sound like (monophtongized and reduced) u. N.B.: as an anaptyctic vowel, ǝ is never stressed. 3

See “Bukaṛa-syndrome” in EALL 2006 (Vol. I).

 

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sina

65

The texts The following text was recorded in Wādiy Ṣlāf4, which lies to the south of Wādiy Fēṛān. The tribe living there, among whom the recordings were made, are known as Awlād Saʿīd. The tribes of southern Sinai are commonly also known as the Ṭawara tribes,5 i.e. the tribes inhabiting the mountainous southern part of Sinai, which is the region often referred to as aṭ-Ṭūr. Together with the dialects of the tribes Ḥamāḏ̣ah, Ṣawālḥah, Garāršah and Ǧbāliyyah, the dialect of Awlād Saʿīd was classified as belonging to group VII in De Jong 2011.6 Awlād Saʿīd Text from Awlād Saʿīd: The Lost Camel 1.) (Š) akallímuḳ7 fi lәǧmāl... (ʿŌ) zimān ǧimálḳ ilabyaḏ ̣8 lam9 ṛāḥ.... inṭálag10 min әṛḅāṭuh walla ēš? walla maṭlag mirtiʿ wall-ēš? 2.) (Š) lā la mʿazzib fi ḥittih, әw ǧamalī́11

                                                                                                                        4 The coordinates of the location where the recordings were made are approximately 28.38.35 North, 33.43.52 East, see Google Earth. 5 In De Jong 2011, the name Ṭawara is used in common reference to the tribes Ǧbāliyyah, Awlād Saʿīd, Ṣawālḥah, Garāršah and Ḥamāḏ̣ah (i.e. goup VII), i.e. it does not include ʿLēgāt (group VIII), Baniy Wāṣil or Mzēnah (the latter two group VI). 6 See De Jong 2011: map 88 (in appendix). 7 Like all of the dialects of southern Sinai (i.e. south of the Tīh escarpment), the dialect of Awlād Saʿīd has /ḳ/ and /k/ as separate phonemes, see De Jong 2011: map 36 (in appendix). These phonemes can be isolated in a minimal pair like waláduḳ “your (masc. sg.) son” – waládk “your (fem. sg.) son”. The superscript u used in the transcription here indicates strong velarisation and rounding of the ḳ, as opposed to its absence in k. Its superscript representation should be understood only in this sense, not as a full vowel. When the u is not superscript, it is to be seen either as a full vowel, as in e.g. ʿinduḳ “with you (masc. sg.)” (as opposed to ʿindik “with you (fem. sg.)”), where -uḳ and -ik are to be interpreted as allomorphs of the personal pronominal suffixes -ḳ and -k, or as an anaptyctic vowel resolving (sandhi) clusters. 8 The article is il- in Saʿīdiy, see De Jong 2011: map 25 (in appendix). 9 lam: “when”, like yōm, lōm. See De Jong 2011: 105. 10 inṭálag: the preformatives of the perfect of measures VII and VIII are not stressable in groups VII and VIII, see De Jong 2011: map 62 (in appendix). 11 -ī́: like in most Bedouin dialects of Sinai, the pronominal suffixes (possessive -ī́ and object -nī́) of the 1st ps sg. are stressed. See De Jong 2011: 68.

66

Rudolf de Jong

máša min ʿindī́ ṭilʿ ilǧabal12, әw lamma ṭilʿ ilǧabal maṛṛaḥ...3.) (ʿŌ) dawwart ʿilēh13 ʿād... (Š) w iṣṣubḥ īǧiy14 fāziʿ luh15, adawwir ʿilēh. fa ana fi lmaġarib šuftuh. әw maṛṛaḥ ... (ʿŌ) [maṛraḥ] fi lḥittah-lliy šuftuh fīheʾ16? 4.) (Š) ạ̄, w adawwir fi lḥittah-lliy šuftuh fīheʾ... lamma ṭilʿit ṛūḥī́ w ana badawwir ʿilēh... (N) mā-luš17 ǧuṛṛah timǝš18 miḥḥeʾ? 5.) (Š) mā ǧuṛṛah, malag, ǧabal... (ʿŌ) ēš ilmalag ḏa ʿād? (Š) ḥiǧǟr, waʿaṛ. waḷḷah w an-īǧiy... w uguṣṣ luh, w uguṣṣ... w inn19 māluh ǧuṛṛah... 6.) (ʿŌ) tuguṣṣ yaʿniy timšiy miʿ ǧuṛṛtuh... (Š) ạ̄ buguṣṣ, gāḷuw [ġār20 hū] mintāg21 šiy... әw yīǧiy wāḥid... әmn ilʿaṛab, gāḷ “waḷḷāhiy šufәt ǧimal... fi ʿaṛabīyt ilʿAṛāyšiyyih”... 7.) (N) hāḏa bәykaḏḏib ʿād... (Š) lā, mā bәykaḏḏib ṣiḥḥ ma bәykaḏḏibš. gāḷ “waḷḷāh әw...” w an-īǧiy fāziʿ. w asʿal22... (ʿŌ) wēn saʿalt? (Š) saʿalt... ilʿAṛāyšiy. gāḷ “ma ḥissēt šiy”. 8.) әw baʿadēn w

                                                                                                                        12 ilǧábal: the article is not a stressable unit in Saʿīdiy dialect, see De Jong 2011: map 18 (in appendix). 13 ʿilēh (< ʿalēh): raising of a in open pre-stress syllables is current throughout group VII, see De Jong 2011: map 83 (in appendix). 14 īǧiy:in this dialect, the prefix vowel of the 1st ps. sg. comm. of the imperfect also harmonizes with the base vowel (i.e. the form is not āǧiy). Other examples are udugg “I hit”, ugūm “I get up”, išīl “I take away”, etc. See De Jong 2011: 88–90. Notice also that the speaker here starts using imperfect verb forms in his story (although it is clearly a description of events in the past), while he uses the perfect verb forms to describe activities that took place further back into the past, i.e. before the sequence of events (described in imperfect forms) that are the framework of his narration. 15 luh: the suffixed preposition for the 3rd ps. masc. sg. is -uh (contrast with -ah or -ih in group I). See De Jong 2011: map 34 (in appendix). 16 fīheʾ: the reflex of final *-ā is raised in neutral (i.e. non-velarized) environments to a position ranging between [e] and [ı], often followed by an unreleased glottal catch (indicated here in superscript). 17 mā-luš: see remark in fn 34. 18 timš: an apocopated imperfect form for the 2nd ps. masc. sg. of the verb maša, yimšiy. 19 w inn or win is often used as a presentative to introduce a sudden (unexpected) development or appearance in a story. Also wlin, cf. Blanc 1970: 34 (145), loosely translatable as “and look there…” 20 ġār: (< ġayr) “it must be that…”, see Hopkins 1990. 21 mintāg: in Turbāniy dialect the word would be míntigiy “having disappeared out of sight”. 22 saʿalt: the verb saʿal, yasʿal “ask” has the reflex ʿ for *ʾ.

 

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sina

67

an-īǧiy ṛākib... fi Ramaḏ̣ān kā[n], w an-aǧǧir lī taks... әmn Aḅuw Rdēs, w aṛawwḥ23 ilʿIrīš... w aṛawwiḥ la.... ilḥaǧǧ Ḥummud24... 9.) yōm25 issuḥūṛ... fi llēl (ʿŌ) ilfaǧәr... (Š) aywah... (N) fi Ramaḏ̣ān ilkalām ḏeʾ26... (Š) Ramaḏ̣ān ạ̄... gult “ǧamal ǧāuḳuṃ27” ṣēftuh ṣēftuh28, gāḷ “waḷḷāh mā ǧāne29”. 10.) guḷt “ṭayyib, fa lǧizzāṛ30 illiy nihā31 wēn?” gāḷ “ilǧizzāṛ... әgṛayyib”. waḷḷāh nīǧiy fāzʿīn l ilǧizzāṛ fi ṣṣubәḥ... әṛṛawwiḥ luh. iṣṣubәḥ... әllaggiy32 ʿa lǧizzāṛ. 11.) gāḷ “waḷḷāh iṛṛāʿiy ʿínduḳuṃ, w izzamal fi l... fi lʿizbih, әw ṛāʿīhin ʿíndihin. w ušūfuw–inkān talguw lēuḳuṃ buʿṛān ʿinduh”... 12.) w aṛawwiḥ. bass hū gāḷ “ilǧamal ḏeʾ... mā ǧiš33 nihāniy”... w ilḥāṣlih baʿad ma... ṛawwaḥna w ma ligīniš34 ilǧamal, ligīne ǧamal la xāḷī́ aʿawaṛ. 13.) zimān ištaṛāh ilʿAṛāyšiy... (ʿŌ) kān fi l... šuġәḷ35 Ṭiʿmih... (Š) ạ̄, šuġәḷ Ṭiʿmih itTīhiy... itTīhiy... w a... w algāh... ʿinduh... marbūṭ... fi lʿizbih. w ilḥāṣlih, w an-īǧiy mṯanniy ṯāniy luh. 14.) guḷt “asmaʿ, ilǧimal                                                                                                                         23 The verb ṛawwaḥ, yṛawwiḥ has the meaning of “go” (as opposed to its more specific meaning for ṛawwaḥ, yiṛawwaḥ in Cairene Arabic “go home”). 24 Ḥummud: I was told that this is a name typically found in al-Arish. 25 yōm: often in the meaning of “when, at the time of”. 26 ḏeʾ: for demonstratives see De Jong 2011: 72–73. 27 -ḳum: for final -m in personal pronominal suffixes of the 2nd and 3rd person masc. pl., see De Jong 2011: 67–70 and map 80 (in appendix). 28 ṣēftuh ṣēftuh: summing up its characteristics, is somehow connected to the root w-ṣ-f. 29 mā ǧāne: the negation is usually formed with the compound ma + …š. In more emphatic negations, one may also hear single ma + verb form, often accompanied by, e.g., w Aḷḷāh (waḷḷāh) “by God” or xāliṣ “at all”, see De Jong 2011: 101–102. 30 ǧizzāṛ: raising of pre-stress a in the pattern CaCCāC is current, but optional in neutral environments, see De Jong 2011: map 22 (in appendix). 31 nihā(-niy) is the current adverb for “here” (co-occurring with híniy), see De Jong 2011: map 46 (in appendix). 32 ǝnnaggiy, assimilated < ǝnlaggiy. 33 For the negation ma ǧiš (of ǧi’ “he came”), though much more regularly ma ǧāš in group VII , see De Jong 2011: 72 (where ma ǧiš is excluded; this should be corrected in this sense). Also see following fn. 34 Like in the example in the preceding fn, *-ā is not lengthened (nor stressed) when suffixed with -š of the compound negation. Other examples are: biddni(’) “we want” – ma biddniš “we don’t want”, biddhi(’) “she wants” – ma biddhiš “she doesn’t want”. Similarly, -u of -u(h) is not lengthened: lu(h) “he has”” – mā-luš “he does not have”, ligītuh “you found him” – ma ligītuš “you did not find him”. The latter form contrasts with ma ligītūš “you (pl. masc.) did not find”, which is the negation of ma ligītum/w. For more detail, see De Jong 2011: 70–72. 35 šuġl: the genitive marker, see De Jong 2011: 66.

68

Rudolf de Jong

ḏa ṣēftuh min ṣēftuh... huwwa... izkān... ṣāḥbuh-lliy-yǧībuh... ʿārif nimṛah-zkān nimṛit baṭāgtuh... w áʿarfuh bass...”. 15.) әw zayy ma yguḷ-ḷuḳ36... íštiruh37 minnuh... (ʿŌ) zayy ma tīǧiy... (Š) zayy ma tīǧiy... waḷḷah w әnhāytuh... w-īǧiy māšiy min ʿinduh, w allagiy ʿala Mḥammad ilĠūl...16.) biddī́ wēn? biddī́ lǦūṛah. gāḷuw “fīh ǧamal... ǧābuh lḤikk-ašgaṛ”, әw ṣēftuh ma ṣēftuh fi lōn ilǧamal šuġḷuh... (ʿŌ) ašgar... (Š) ạ̄, w әmʿaḏ̣ḏ̣uṛuh... azaṛag38... min waṛaʾ. 17.) w īǧiy mṛawwiḥ... l ilǦūṛah. yōm ṛawwaḥt ilǦūṛāh ma ligītš... ʿŌdih, ligīt waladuh... 18.) saʿaltuh guḷt “uḅūḳ ǧāb ǧamal šiy ṣēftuh min ṣēftuh” gāḷ “waḷḷāh mā ǧāb”. w illiy mʾaǧǧirīnuh yōm ṛawwaḥna lǦūṛah taḏ̣āyag gāḷ “intuw xaḏtūnī́ l ilʿIrīš, әw... ilḥīnih... 19.) (ʿŌ) biddḳuṃ ilǦūṛah... (Š) biddḳuṃ ilǦūṛah”. guḷt “yā ṛāǧil, inta la taḏ̣āyag39 wala ḥāǧih... binḥāsíbuḳ inta btímәš b ilkīluh. ’iǧāṛuḳ maʿaṛūf... illiy mn Aḅuw Rdēs... ēh? 20.) (ʿŌ) mīn ímʿuḳ40 ṛafagāuḳ? (Š) imʿī́ Ḥsēn ṛaḥamat әḤsēn imʿī́. (ʿŌ) әḤsēn Aḅuw ʿAṭwa... 21.) (Š) ġēr ana w hū tinēnneʾ41. gulna “-ǧāṛuḳ tạ̄xḏuh-nta. hū māšiy b ilkīluh w inta btimš, wala tazʿal wala ḥāǧih”. 22.) waḷḷah әw ṛawwaḥna lǦūṛah, tigahwēna42 ʿind irriǧǧāl әw saʿalnāh... wala ... mā ḥadd áṭra lǧamal aṣliy43 māḥadd áṭra...23.) w īǧiy әmṯannyīn ṯāniy, w әṛṛawwiḥ l Aḅuw Ǧirēr... (ʿŌ) fi lʿIrīš... (Š) fi lʿIrīš. guḷna.... ilǧimal ṣēftuh min ṣēftuh, “ʿāyzuḳ túšuf44 lī...”. 24.) gāḷ “úgәʿd-úgәʿduw úgәʿduw... úgәʿduw”. әw kān miʿne... ilḥaǧǧ Ḥummud kumān... 25.) “náʿaml ēh ḏ-ímʿuh?” gāḷ “asmaʿ... yā waladī́” gāḷ “ilǧamal... uṭlub ʿawaḏ̣uh min ṛabbna bass asʿal ʿinnuh... uṭlub ʿawaḏ̣uh min ṛabbna bass asʿal                                                                                                                         36 For suffixed prepositions, see De Jong 2011: 75–77. 37 íštiruh: an apocopated imperative (of tertiae infirmae verb ištara, yištiriy) followed by the object suffix: íštir + -uh. See also tim(ә)š in sentences 19 and 21. 38 azaṛag (a bukaṛa-form): lit. “blue”, but in the colour system of Bedouin it is often used for other dark colours as well, like dark brown, or (euphemistically) black. 39 la taḏ̣āyag “don’t be angry”, with reduced initial tt cluster of this measure VI verb (< * la ttaḏ̣āyag). See De Jong 2011: 96–97. 40 ímʿuḳ: for suffixed prepositions, see De Jong 2011: 75–77. 41 tinēnneʾ “the two of us”: notice the reflex t for *ṯ. 42 The verb tagahwa, ytagahwa can mean “drink coffee” or “drink tea”. 43 I could not find a meaning for áṭra, but I was told that in Turḅāniy dialect this would be something like: mā ḥadd ǧāb ṭiryātah b ilmaṛṛah “Nobody had any news about it at all”. 44 túšuf (instead of tšūf): in hollow verbs (i.e. second radical = y or w), the 2nd ps masc. sg. often has a reduced long vowel of the verbal stem (as a result of which stress also shifts). Another example I heard is (when I was asked after taking some photos): hatíğib áṣṣwar, walla tánam ʿilēhin? “are you going to bring the photos or are you going to keep them?” (instead of the – also possible – forms tǧīb and tnām). See De Jong 2011: 84-85.

 

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sina

69

ʿinnuh”. 26.) guḷt “ṛabbna kulluh ʿawaḏ̣”. waḷḷah w ilḥāṣlih... wiṣṣēt ʿād ilǧizzāṛ... adṛaktuh, guḷt “in ǧāuḳ ilǧamal...” zayy ma guḷt luh ʿād. 27.) waḷḷah әw ǧīneʾ... әw salāmítuḳ w irdabb ʿāfyih! (ʿŌ) lākin fīh ḥāǧih... ma ligītuš45 inta. ligītuh ṭabbēt fīh? ma ʿallamtǝš ṭabbēt fīh kēf... 28.) (Š) lēh? biddḳuṃ uṭubbuh fīh? (ʿŌ) ạ̄... (Š) ū:hu! (ʿŌ) ilǧamal ʿāwzīn náʿarfuh wēn... [laughter] [...] әnhāytuh nīǧiy mn ilʿIrīš... әw nīǧiy fāzʿīn... ítla blād ilәMzēnih, w әndawwir fi blād ilәMzēnah, fi lḤargūṣ әw fi... Uṃṃ Buwāṛig әw miš ʿārif әw... әw fi lʿIšš... әw fi... 30.) ilḥāṣlah ma xiḷḷēniš46 maṭāriḥ ma ǧīnāhiš47. әw nīǧiy, әw nafzaʿ ṯāniy... әndawwr ilǧamal. w әnlaggiy wēn w әnlāgiy wāḥid šuwwāf, huwwa-lliy xaḷḷāna nadawwir nihā. 31.) gāḷ “ilǧamal... ʿind wāḥid... әw mustaḥill luh. әw byasṛaḥ әw byiḏ̣wiy miʿ ǧamaluh”. 32.) guḷt “әbyasṛaḥ w әbyiḏ̣wiy48 miʿ ǧamaluh әw hū mistaḥill luh nāwiy... nāwiy ʿilēh... yạ̄xḏuh”. 33.) әw nafzaʿ, naṛkab ṯāniy... ṛakәb. w әnṭubb isSʿāl... w ilfarš49 min dūn isSʿāl... әmnaṭlaʿuh... b irriǧәl. 34.) xillēna buʿṛānna fi ʿirgah w ṭiliʿne. ana w Xiḏ̣r Aḅuw Sālim, әw nnn... naḥarid Zaġaṛah, w әndawwir әNmēr. 35.) әndawwir ilǧamal fi lmaṭāṛiḥ illiy fīha ṃayyih. w әnguṣṣ inkān ǧuṛṛt ilbuʿṛān fīhe. w әndawwir әNmēr kumān. 36.) w әndawwir әNmēr әw linn ǧuṛṛit ǧamal hīhī giddāmna ḥimṛāʾ50, bәysannid әw biysīḥ әl waḥaduh... 37.) (ʿŌ) guḷt “hāhū ḏeʾ” (Š) guḷna “hāḏa hū”. fi Rās Šuġḷ ilAṣḥab, әw hū ḥāṭṭ fi talʿah min táḥatuh, min taḥat ǧamaluh. 38.) әw nīǧiy, w әṛṛawwiḥ yā-ḅuw ṣḥayybī́ [naṣṭabīhin] min әNmēr, әw min әNmēr әnlaggiy Sanad, әw fōg fi Sanad farš... 39.) (ʿŌ) ṛakәb? (Š) ṛakәb. әw min Sanad ... ṛawwaḥna fi lʿAṛāynih, әw ṭabb ʿilēna maṭar. w әnmarriḥ fi l.. fi lʿArāynih. 40.) w iṣṣubәḥ min ʿArāynah ǧīna kāttīn, w әnkutt ʿa ṭṬarfa, әw minnuh nīǧiy ʿaṛabne. 41.) әw yōm ǧīna... fazaʿ zalamih... bәydawwir luh biʿīr. gāḷ “waḷḷah ana ligīt biʿīr”... ṣēftuh min ṣēftuh... (ʿŌ) mayyit... (Š) mayyit. 42.) guḷna “wēn?” gāḷ “fi ḥittah lflāniyyih” w īǧiy fāzʿ ímʿuh... iṣṣubәḥ. 43.) (ʿŌ) ʿind ilmalag... (Š) ʿind ilmalag, fi farәš... guḷt “yāh... Aḷḷāh” guḷt “ḏ̣aṛṛítuḳ fi ḏ̣aṛṛítuḳ hāḏa ǧamalī́ nībānuh әṭwāḷ ka hinna w šaʿaṛuh hēhū? 44.) әw šibhuh w ṛāsuh gbāluḳ hēhū...” gāḷ “waḷḷāh miʿ bālī́ ǧamáluḳ.” hēhū nafsuh... (N) mayyit ʿind ilmalag. 45.) (Š) mayyit fi ʿirg ilmalag míntišib luh fi ḥaǧar, әw míngilib әw mínkisir. ʿirәg                                                                                                                         45 ma ligītuš: see remarks in fn 34. 46 ma xiḷḷēniš: see remarks in fn 34. 47 ma ǧīnāhiš : see remarks in fn 34. 48 ḏ̣awa, yiḏ̣wiy “return home at sunset (with small cattle)”. In group I dialects of Sinai this is often a measure IV verb: aḏ̣wa, yiḏ̣wiy. The verb seems somehow connected with the Classical Arabic root ḍ-w-ʾ “light”; using the last light of the day to go home. 49 farš: a plain covered with small rocks, with a surface area smaller than a ḥamādah. 50 ḥimṛāʾ: see De Jong 2011: 63.

70

Rudolf de Jong

ṛamḷah. (ʿŌ) zixәb (Š) ạ̄, zixәb fi ʿirg ilmalagah... ʿirәg malagah. 46.) (ʿĪ) yabbas rīgak ǝw hū ʿind máṭraḥak. (Š) әw hū fi maṭraḥuh. 47.) iḥna bәndawwir bәntaʿadda ʿilēh min fōg, әw mi nhā mi nhā, әw hū malbūg ma bēn әḥṣíy, әw bēn ilmalagah. bēn izzuxbah. 48.) (ʿŌ) inta ʿārif ilmalbūg yaʿniy ēh? (R) laʾ. (ʿĪ) ilmalbūg hēḏiy hayṭalliʿ rūḥī́ bukṛah... [laughter] (Š) әyṭalliʿ ṛūḥuḳ51 ʿilēha? 49.) (ʿĪ) ạ̄ ġār atarǧimha lah halḥīn. lā, bass hū biʿīr ġāliy ʿindak, әtdawwir ʿaláh. (Š) ạ̄ hū biʿī́r ġạ̄liy ʿindī́, biʿīr ḥilәw. (ʿĪ) ḥatta lamma ligītah mayyit irtaḥt... (Š) la xalāṣ min, yōm ligītuh... (ʿŌ) ilgiṣṣah ḏiy ḥigīgah ḥaṣalát-luh hū... Translation: The Lost Camel 1.) (Š) I’ll talk to you about camels… (ʿŌ) In the old times, when your white camel disappeared… Did it come free from its rope or what? 2.) (Š) No, no, I was camping at a place, and my camel went away from me and went up the mountains, and when it went up the mountains, rested… 3.) (ʿŌ) So you went looking for it… (Š) And the next morning I went looking for it, (and) I looked for it. Because I had seen it at sunset. And it was resting… (ʿŌ) [Was it resting] at the place where you had seen it? 4.) (Š) And I looked at the place where I had seen it. Until I went crazy looking for it… (N) Didn’t it have tracks that you could follow? 5.) (Š) No tracks, hard rock, mountains (ʿŌ) So what is this malag? (Š) Rock, difficult terrain. By God, and as I was coming… and I was following his tracks, and I looked… there (suddenly), it hadn’t left tracks… 6.) (ʿŌ) You were following his tracks… (Š) Yes, I was following (his) tracks, they said [it must] have gone out of sight, or something… and then someone came of the clan/family (who lived there), he said “By God, I saw a camel in the car of the ʿAṛāyšiy people… 7.) (N) So this man was lying… (Š) No, he wasn’t lying, it was true, he wasn’t lying. He said “By God and…” And I got worried. And I asked… (ʿŌ) Where did you ask? (Š) I asked… the man from al-Arish. He said “I haven’t noticed anything”. 8.) And after that I came riding… it was during Ramadan, and I rented a taxi for myself… from Abuw Rudeis. And I went to alArish… and I went to… Ḥajj Ḥummud… 9.) At the time of the meal before sunrise, at night. (ʿŌ) At daybreak… (Š) Yes… (N) This was during Ramadan… (Š) Ramadan, yes… I said “Did a camel come to you?” I described it, he said: By God, it didn’t come to us”. 10.) I said “So the butcher who is here, where is he (i.e. where does he live)?” He said “The butcher… is near”. By God, we came                                                                                                                         51 ṛūḥuḳ: the vowel sounds like a full u, but it is best interpreted as an anaptyctic vowel resolving the sandhi cluster ḥḳ ʿ here.

 

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sina

71

looking for the butcher in the morning… we went to him… we went to the butcher. 11.) He said “By God, (I want to see) the herdsman with you, and the camels at the… at the farm, and the herdsman with them, and I (want to) see whether you find camels with him that belong to you (pl.)”… 12.) And I went. But he said “This camel… didn’t come here” … And what happened after… we went (there) and didn’t find the camel, we found a half blind camel that belonged to my uncle. 13.) The man from al-Arish had bought it long time ago… (ʿŌ) It was in… it belonged to Ṭiʿmih… (Š) Yes, the Tīhiy camel that belonged to Ṭiʿmih… the Tīhiy (camel)… and I… and I found it… with him.. tied… at the farm. And what happened, and I went back again to him. 14.) I said “Listen. This camel (and I described) its characteristics… he… if… its owner who brings him… (I want to) know the number, if the number of his ID card… I just want to know it (so I’ll know who he is)…”. 15.) And whatever he says to you… buy it from him… (ʿŌ) Whatever comes (i.e. whatever he asks for) (Š) Whatever comes… By God, and the end of it... And I came walking back from him, and I went to Mḥammad ilĠūl.. 16.) Where did I want to go? I wanted to go to ilǦūṛah52. (Because) They said “There is a light coloured camel that ilḤikkiy brought, and they described the colour of his camel… (ʿŌ) blond53 … (Š) Yes, and its back dark brown… from behind. 17.) And I came going… to ilǦūṛah. When I went to ilǦūṛah, I did not find… ʿŌdih (i.e. ilḤikkiy), I found his son… 18.) I asked him (and) said “Did your father perhaps bring a camel which looked like such and such?”... he said “By God, he didn’t bring (one)”. And the man whom we had hired (i.e. the driver of the taxi) became angry when we went to ilǦūṛah. He said “You have taken me (all the way) to al-Arish and… now… 19.) (ʿŌ) you want to go to ilǦūṛah… (Š) You want to go to ilǦūṛah”. I said “Oh man, don’t be angry or anything… we settle our account with you as you go per kilometre. Your price is known (to us)… which is from Abuw Rudeis… what?” 20.) (ʿŌ) Who were your travelling companions (that were) with you? (Š) Ḥsēn was with me, the late Ḥsēn was with me. (ʿŌ) Ḥsēn Aḅuw ʿAṭwa... 21.) (Š) Only I and he, the two of us. We said “You will get your fee. It is (calculated) per kilometre as you go, don’t be angry or anything…” 22.) By God, and we went to ilǦūṛah, and we had a hot drink with the man and we asked him… not… nobody had any news at all from the camel… Nobody had                                                                                                                         52 ilǦūṛah is about 30 kilometres to the east-southeast of Al-Arish. 53 “blond” is a too literal translation. The colour is a light yellowish brown. See also Procházka in Online EALL, lemma “Color Terms”.

72

Rudolf de Jong

any news… 23.) And I came back again, and I went to Abuw Ǧrēr… ʿŌ) In alArish... (Š) In al-Arish. We said… “The camel with such and such characteristics, I want you to see for me…” . 24.) He said “Sit down, sit down, sit down… sit down.” And with me was… Hajj Ḥummud also… “What are we going to do with him?” 25.) He said “What are we going to do with him?” He said “Listen, my son” He said “The camel… just ask for it to be compensated by our Lord, ask (to receive) in return of it… Just ask for it to be compensated by our Lord, ask (to receive) in return of it…” 26.) I said “Our Lord is all compensation (one needs)”. By God, and what happened… so I impressed on the butcher… I made it very clear to him, I said “If the camel comes to you…” so like I had said to him (i.e. to ʿŌdah). 27.) By God, and we came… and (I greeted him) “Peace upon you and heaps54 of strength!” (ʿŌ) But there is one thing… you hadn’t found it. Did you find it, did you find it? You haven’t told how you found it… 28.) (Š) Why? Do you want me to find it? (ʿŌ) Yes… (Š) Oooh! (ʿŌ) The camel, we want to know where it was… (laughter) […] 29.) The end (of the story) was that we came (back) from al-Arish… and we came looking… in the direction of the land of the Mzēnih, and we searched in the land of the Mzēnih, in ilḤargūṣ and in... Uṃṃ Bawāṛig and I don’t know, and… and in ilʿIšš... and in... 30.) What happened is we did not leave (any) places where we didn’t come. And we came, and we searched again… and we looked for the camel. And we went where? And we found a seer, he is the one who made us look here. 31.) He said “The camel… is with someone… who uses it as he deems fit. And he goes grazing and comes back home with his camel.” 32.) I said “He goes grazing and comes back in the evening with his camel, and does with it as he pleases… whatever he likes (to do) with it… he takes it”. 33.) And we went searching, we mounted again… riding (on camels). And we went to isSʿāl… and the rocky plain beyond isSʿāl. We went up there… on foot. 34.) We left our camels behind in a narrow wadi and went up. I and Xiḏ̣r Aḅuw Sālim, and we crossed through (Wadi) Zaġaṛah55, and we searched (in) Nmēr.56 35.) We looked for the camel in places where there is water, and we followed tracks when there were camel tracks. And we also looked in Nmēr. 36.) And as we were looking in Nmēr, suddenly there were the tracks of a camel and there was a red she-camel in                                                                                                                         54 Lit. “an ardeb of strength”; an ardeb is a dry measure (Egyptian) of 198 litres, cf. Wehr 1980. 55 There are several (narrow) wadis named Zaġarah. 56 Nmēr is a location in Wadi Zaġarah.

 

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sina

73

front of us, he was going up and roaming about by himself… 37.) (ʿŌ) I said “There he is”. (Š) We said “This is him”. In Rās Šuġḷ ilAṣḥab, and he had made camp in a flat area in the (rising) wadi, downstream from it, downstream from his camel. 38.) And we came, and we went, oh my little friend57 [we looked for them??] from Nmēr, and from Nmēr we went to Sanad, and up in Sanad to a rocky plain… 39.) (ʿŌ) (were you) Riding (camels)? (Š) (we were) Riding (camels). And from Sanad… we went in ilʿAṛāynih, and (then) we got caught in a rain shower. And we slept in ilʿAṛāynih. 40.) And in the morning from ʿAṛāynih we came going down the wadi, and we went down the wadi to aṭṬarfa, and after that we came to our (own) people. 41.) And when we came… a man was searching… was looking for a camel. He said “By God, I found a camel”… with such and such characteristics… (ʿŌ) Dead… (Š) Dead. 42.) We said “Where?” He said “In the place so-and so”. And I went looking with him… in the morning. 43.) (ʿŌ) Near the hard rock (area)… (Š) Near the hard rock (area), in a rocky plain… I said “Oooh… God”, I said “Your fart upon your fart!, this is my camel.” Its teeth were long like them (i.e. the teeth of my camel) and its hair, there he was! 44.) And its likeness and its head in front of you, it was him… He said “By God, I think it’s your camel.” It is the very same… (N) It (was lying) dead in the hard rock (area). 45.) (Š) (Lying) dead in a fissure in the hard rock (after) having tripped over rocks, and it had turned (on its back) and broken (i.e. it had broken bones). A sandy fissure (ʿŌ) A narrow passage (Š) Yes, a narrow passage in the canyon of the hard rock… a canyon in hard rock. 46.) (ʿĪ) He had worn you out (lit. he had made your saliva go dry), while he was (actually) near your place (all that time) (Š) While he was in his place. 47.) We were searching and we came past it from above, and this way and along here, while it lay hidden between rocks, and in (lit. between) the hard rock. In (lit. between) the narrow passage. 48.) (ʿŌ) Do you know what malbūg means? (R) No. (ʿĪ) This malbūg he’ll drive me crazy about it tomorrow… [laughter] (Š) Will he drive you crazy about it? 49.) (ʿĪ) Yes, I have to translate it for him now. No, but this camel was dear to you, that you went looking for it. (Š) Yes, the camel was dear to me, it was a beautiful camel. 50.) (ʿĪ) Even so that you found rest when you had found it dead. (Š) No, that’s it, from the moment I found it… (ʿŌ) This story is the truth, it (really) happened to him…

                                                                                                                        57 Lit. “Oh father of my little friend”.

74

Rudolf de Jong

Tayāha The following texts were recorded from the Tayāha in the area of al-Axaḏ̣ar (to the north of Wadi aṭ-Ṭarfa, just south of the Tīh escarpment). Tīhiy Arabic is a group I dialect and a description of this dialect may be found in De Jong 2011: chapter 3. Text from the Tayāha: The Accident 1.) (ʿĪ) alhawyah58 hēḏiy kēf? kēf ṭabbat fak59 yā Mḥammad? (M) hēḏiy kunt maḥabūs60, ṭilíʿәt sittih w tisʿīn, fa ǟh? ṛawwaḥt aššimāl әhnúh61… 2.) w Aḷḷāh әw ḏ̣aḷḷayt62.. ǧibәt ʿaṛabiyyah ʿād aná... b aṛbaʿah w ʿišrīn... b aṛbaʿah w ṯalāṯīn alf. kānat alʿamaliyyih māšyih fi sittah w tisʿīn. 3.) w aǧīb alʿaṛabiyyih w ana mṛawwiḥ... w alʿaṛabiyyih yimkin ǟh? miyyit mitǝr ʿan albēt... gult l arraǧǧāl ʿād naṣībak bәyǧī-bak. 4.) w arriǧǧāl w aná, w iḥna ǧīna ndawwir, yōm dawwaṛna lʿaṛabiyyih, gāl widd-asūg 63 ʿaṛābīytak, nizilt әw rikíbәt min baṛṛa. 5.) iḥna ṯalāṯ t-infāṛ. rikíbәt64 wēn? albāb albuṛṛāniy... ǧāl aṛṛākib. w Aḷḷah w aṛkab... yōm agbalna ʿa lbēt... fīh ġarzih...6.) basūg w ana ṣġāyyir zamān... gult inta ʿārf әtsūg? 7.) gāḷ “basūg fi lġarzah әlhá swāgih”. w Aḷḷāh lamma xašš alġarzih... 8.) min ṯāliṯ yaʿniy miš māšyih ʿázәmhiy, ṛadd әlhiy ṯāniy... laffat minnih alʿaṛabiyyih. 9.) (ʿĪ) saffaṭ… (M) ạ̄ saffaṭ saff kiḏiy... yōm saffaṭ... álʿaǧal 65 ṭiliʿ ʿan álmadag baṛṛa... algiddāmāniy... 10.) ġāṛ win66 ḏ̣abḥithiy. arriǧāl                                                                                                                         58 alhawyah: “the accident”, notice that the article is al- in Tīhiy. See De Jong 2011: map 25 (in appendix). 59 fak: lit. “in you”; in Tuṛbāniy, the paradigm of the preposition fi with sg. pronominal suffixes is: fah / fīha; fak / fīkiy; fay(y) “(respectively) in him/her; in you (masc./ fem.); in me”. For suffixed prepositions, see also De Jong 2011: 245–248. 60 maḥabūs: after insertion of a due to the gahawah-syndrome (< maḥbūs), see EALL, lemma “Gahawah-Syndrome”. 61 Like in most group I dialects, the adverb for “there”is hnuh, see De Jong 2011: map 45 (in appendix). 62 ḏ̣aḷḷayt: in velarized environments and following ḥ, ʿ, x and ġ diphthongs ay and aw tend to be preserved, see De Jong 2011: map 11 (in appendix). In otherwise neutral environments, reflexes of *ay and *aw are usually ē and ō, but see e.g. also kattayt “I went downstream in the wadi” in sentence 30. See also remarks in De Jong 2000: 77–80. 63 widd-asūg: widdī́ + asūg. 64 In the perfect of verb forms reflecting *faʿila, the unstressed i of the first syllable is not dropped, see De Jong: map 52 (in appendix). 65 álʿaǧal: the article is al- and is stressable in strings like al-CvC(vC), see De Jong 2011: map 18 (in appendix).

 

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sina

75

tawaffa fi maṭraḥah, w anáʾ... riǧlāy aṯṯintēn īdak ʿalēhin mín-ihniy67. әw ǧāmb ʿaynī́ hēḏiy, w īdī́ hēḏiy. 11.) w Aḷḷāh w gaʿadt fi lmistašfiy xamsah w aṛbaʿīn yōm... ʿugәḅ alxamsah w arbaʿīn. 12.) (ʿĪ) fi yāt mistašfiy? (M) mistašf-alʿIrīš... almistašf-alʿāmm hēḏiy. w Aḷḷah w ugʿud xamsah w arbaʿīn yōm w an-aṭlaʿ... 13.) әw minnih ʿād ēh? yōm ṭiliʿt sawwēt әwṛāgī́, mā mʿay baṭāgah alhawiyyih hāḏ̣āk alḥīn sawwēt alhawiyyih, әw gaddamt, 14.) gāl әnǧīblak kaṛṛūsih walla ǧhāzāt? gult laʾ әnšūf alәǧhāzāt... law kaṛṛūsah kān lagētnī́ l alḥīn alʿamaliyyah taʿbānih. 15.) w Aḷḷah w aǧīb alәǧhāzāt... ramēthin šaháṛ f-addāṛ, lamma taʿawwadt ʿālēhin baʿadēn. 16.) taʿawwadt ʿālēhin alḥīn basūg, әw baṛkab ʿa lbiʿīr... yaʿniy kān ēh? alәmḥayyirnī́ , masalan alǧarәy mā ḥayltī́ aǧiriy ǧiyyid, әw mašәy sirīʿ. amma bāgiy lḥaṛakāt tamām... ṭibīʿiy... 17.) (ʿĪ) hāḏ̣a min әmxaḷḷafāt aḷḷġūm... (M) ǧēš iyyām sabʿah w sittīn... (ʿĪ) bass ʿārif almanṭigah hēḏiy fīha lġūm... (M) mā-driy ʿanhiy ana dāriy ʿanhiy inn fīh әḷġūm walla mā fīh... 18.) madagg, zamān assēl ǧāyibhin min fōg... maʿ alblād maʿ aṛṛamyih hēḏiy ǧāyibhin aḷḷġūm... әmsawwyih ʿād ēh? ṛamyih ṛamәḷ ʿalēh68 min fōg. 19.) ṭabʿan, min aṛṛamәḷ hāḏ̣a... әbyasfaḥah aṛṛīḥ byasfaḥah ǟš? min bәygarrib bәyṛaffḥ alblād aṭṭabagah... aṭṭabagah lliy bēn aḷḷuġәm bәyṣīr... bәyṣīr әṛḥayyfih xāliṣ... 20.) әtṣīr lamma... lōm69 kibīrih, mā bәtḥīg fīh alʿaṛabiyyih... Text from the Tayāha: The Mother of Xiḏ̣r 1.) (ʿĪ) intiy Aṃṃ Xiḏ̣әr? (A) ayw-Aḷḷah! w Aḷḷ-Aṃṃ Xiḏ̣әr rabbna yxalliy… ywarrīnī́ wlēdī́70... Aḷḷāh... әyxalliy lih әwlēdātih... (Ḏ̣ayfaḷḷah) [...] (A) [to her son] wǟš wǟš? 2.) (ʿĪ) yā-ṃṃ Xiḏ̣әr intiy min wēn... aṣlan? (A) aṣlī́ mn alʿAwāmrih, annās aṭṭayybīn bass mātow Aḷḷāh yaṛḥamhuṃ... (ʿĪ) Aḷḷāh yaṛḥamhuṃ... 3.) (A) Aḷḷāh yaṛḥam alǧimīʿ, bass... (A) atīm-anáʾ atīmih.. (ʿĪ) bass sukkān... (A) aḥna kullna ṯalāṯih... Wardih, waḥdih mgaṣṣirih fi Nixǝl, әw wāḥid ʿind әMzēnih... w aḥna...4.) w aná taǧawwazt ʿind zalámah zēn Mzēniy markad alǧannih, min lәĠṣayynāt min Ḏaháb nās zēnīn. 5.) māt aḅuw lʿayyil hāḏa w hū ṣġāyyir. әw hāḏ-anáʾ... [w ataṛazzag b Aḷḷāh] әw taʿazzáṛәt әw fi ṯṯalǧ... anguḷ lamma ṛṛṭūbah katalatnī́ halḥīn. 6.) әw ʿišt, w ilḥamdillāh f-álәxšaš                                                                                                                                                                                                                                                                                                                                                             66 win or wlin may be used to introduce a sudden and unexpected development in a story, see Blanc 1970: 34 (145). 67 mín-ihniy: min + hniy forms one single stressable unit. 68 ʿalēh: raising of a in such positions, as in ʿilēh, is actually more current in Tīhiy, see De Jong 2011: 245-246. 69 lōm: “when” (also yōm), see Blanc 1970: 35 (146), De Jong 2000: 230 and ibid. 2011: map 71 (in appendix). 70 wlēdī́: like most Bedouin dialects in Sinai, the possessive (-ī́) and object (-nī́) suffixes of the 1st ps comm. sg. are stressed.

76

Rudolf de Jong

hāḏ̣aḷ... әw aṛbaʿ әl fōg w aṛbaʿ әl taḥát w alfaṣәl gaṛáʿ annās әhnúh... ʿannī́, w anguḷ alʿayš әmn alwādiy hniy... w álmiy hniy. 7.) w álḥaṭab bass álḥaṭab zamān wāǧid... әw yōm gaṣṣarna xalāṣ әmn álwaǧaʿ... 8.) lōla lʿayyil әwlēdī́ byaxadimnī́ [...] (ʿĪ) bass intiy min sukkān wēn... (A) [w Aḷḷah] min yōm xilígәt71... f-alәblād hēḏiy... f-alĠaṛábah әw fi Fēṛān... f-álĠaṛábah w fi Fēṛān. әw hāḏ-anáʾ... 9.) min gaḅәḷ ḥarәb... min gaḅәḷ... gaḅәḷ Ǧamā:::l әymūt... (ʿĪ) Ǧamāl aṛṛayyis šuġәḷ Maṣәr zamān... (A) Aḷḷāh yaṛḥamih Ǧamā:::l ʿAbd inNāṣir... (ʿĪ) ạ̄... 10.) (A) anáʾ fi Fēṛān, әw ǧīt aṭṬaṛfa. әw halḥ... әw ʿugәḅ sanatēn gām alḥarәb... (ʿĪ) ʿugәḅ sanatēn baʿad alḥarәb... (A) [...] ʿugәb sanatēn. (ʿĪ) hāḏ̣a ḥarәb sabʿah w sittīn?(A) barḏ̣aʿ әHlālih... barḏ̣aʿ әHlālih, әw ǧibәt Sālim fi Fēṛān. әw baʿád ṯalāṯ әsnīn axaḏnī́ aššēbih Sillām. 12.) әw yōm ǧibәt Xiḏ̣әr... assāʿ mā barḏ̣aʿ gām ḥarәb Isṛāʾīl ʿalēna... ʿām sabʿah w sittīn, fi hāḏ̣a72 f-aṭṬarfa... (ʿĪ) ḥarәb sabʿah w sittīn... (A) w aná f-alblād hēḏiy xāliṣ... mā ṭilíʿәt minhiy b ilmaṛṛah. kān lay ʿamm bәyǧīnī́ Aḷḷāh yaṛḥamih yōm tawaffa hāḏ-anáʾ... fi makānī́... әw lih әʿyāl, w annās... w alәʿyāl azzēnīn mātaw... 14.) (ʿĪ) әw sakantiy... (A) ġayr hāḏa misāknī́. (ʿĪ) әw kān kēf Fēṛān әyyām ma ǧītiy hniy? (A) w Aḷḷah mā ǧītih gidīm. ǧibәt Xiḏ̣әr әw Ḏ̣ayfaḷḷah w ana mā ǧītih... ǧibthuṃ fi lәblād hēḏiy...15.) (ʿĪ) fi lәblād mā kattētiy? (A) lā la la la la. ǧibt fi hāḏa Xiḏ̣әr fi ṭṬarfa, әw Ḏ̣ayfaḷḷah waṛa lǧidlih hēḏīk...16.) әw hāḏanáʾ f-alblād hēḏiy... ʿind ilǦbāliyyih mā futt... (ʿĪ) ạ̄... (A) alәǦbāliyyih.... (ʿĪ) ṭayyib... (A) alәǦbāliyyih nās әkwayysīn, lā bәyrīduw mišākil... 17.) (ʿĪ) ṭayyib xarrfīnī́... ana widdī́-yyākiy txarrfīnī́ kēf әtsawwiy ssamәn aššīḥiy... (A) assamn aššīḥiy ḥilәw. (ʿĪ) ạ̄... (A) yā ṛabb yurzugna b alḥalāl... (ʿĪ) yā ṛabb... bass kēf bәysawwūh? 18.) (A) bәysawwūh, bәyxuḏ̣ḏ̣ūh... walla widdak xaḏ̣ḏ̣ walla widdak әtḥa... (ʿĪ) min yōmin ma

ḥalabawh... (A) yōm yaḥalbūh bәyṛawwgūh... 19.) (ʿĪ) ạ̄... (A) w әyxuḏ̣ḏ̣ūh, w әyzabbdūh... (R) fi ssiʿǝn... (A) әw yilәḥgūh ʿa nnāṛ... ạ̄ yxuḏ̣ḏ̣ūh fi ssiʿәn... w әyzabbdūh. 20.) yōm yukṯur azzibdih ymayyʿūh... ʿa nnāṛ, әyḥuṭṭuw ʿalēhiy milḥ әw kuṛkum... әyṣaffūh, әw yōm yubṛud әyḥuṭṭūh... 20.) (ʿĪ) әyḥuṭṭuw ʿalēh ziṛagṛag? (A) äh? (ʿĪ) bәyḥuṭṭuw ʿalēha ziṛagṛag? (A) lā, әyḥuṭṭuw ʿalēhiy kuṛkum ’aṣfaṛ... (ʿĪ) kuṛkum ’aṣfaṛ...21.) (A) aywah zamān w int әbtarḏ̣aʿ... kān ʿindina miʿziy, kammalat-. [...] ạ̄, kuṛkum ’aṣfar, yagdaḥawh maʿ almilәḥ fīh... 22.) әw minnih ǟh? әysawwlūh yōm yabṛa::d… fi maʿūn naḏ̣īf... (ʿĪ) ṭabʿan әykūn maʿūn naḏ̣īf... 23.) (A) ạ̄ hāḏa swāt alkuṛkum, hāḏa ssamn aššīḥiy... (ʿĪ) hāḏ̣a ssamәn... (A) assamәn hāḏa mā bәnḏūgih, ism Aḷḷah ʿaláy min ṣadrī́, lā-ḏūg samәn wala zēt. 24.) әmn aṯṯalǧ w albard alliy šuftih zamān... ism Aḷḷah ʿaláy ʿīšt alәʿyāl b adayy... (ʿĪ) ʿīšt ēš? (A) ʿīšt әḏ̣ʿūfī́ b adayy,                                                                                                                         71 A very rare example of an internal passive form, cf. Classical Arabic *xuliqtu. In poetry, internal passives are less rare. 72 fi hāḏ̣a “here”, cf. Stewart 1990: 187 (text 69), line 216.

 

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sina

77

albint... (ʿĪ) ạ̄... 25.) (A) әw Xiḏ̣әr әw Ḏ̣ayfaḷḷah hāḏa f-álәxšaš hāḏa ṛaḅḅaythuṃ b adayy. baḥaṭṭib maʿ annās w aǧībih... ʿīših w alḥamdillāh... (ʿĪ) álḥaṭab hāḏa ēš nōʿ... (A) álḥaṭab mā ḥadd lāgīh l annāṛ, katalna lbarәd... 26.) (ʿĪ) bass fi lblād hēḏiy b nōʿ álḥaṭab ēš? (A) hāḏa ʿaǧaram. әbnayytī́ lbāriḥ gōṭarat, naggatih lay w aná wiǧʿānih. 27.) ʿaǧaram f-alblād bass mā byínligiy bass kammalawh. annās ġāz w aḥna mā lna ḥīlih nǧīb alġāz... (ʿĪ) maḥál, maḥál... álmaḥal ḏabáḥ alblād. 28.) (A) w Aḷḷāhiy ʿinna gaṭáʿ alblād yā ʿĪd. ḥatta lbarәd katalna f-allēl... zamān albʿayṯaṛān. (ʿĪ) aywah... (A) әw ratám... w әwṛāgah, әw šīḥ, әw ʿaǧaram, әw kull šiǧáṛah... 29.) (ʿĪ) xayṛāt Aḷḷah... (A) falәǧbāl ḏ̣aḷḷāk hēḏiy w aná mikīn-asṛaḥ. kammal áššiǧaṛ w alḥalāl әw [...]. 30.) (A) ʿašaṛ әsnīn alwādiy mā kattaytih... alwādiy hāḏa xamәsṭāšaṛ sanáh mā katteytih. hāḏ-aná yā ʿĪd ʿašaṛ t-әyyām w ana wiǧʿānih [coughs] wiǧʿānih... [clears her throat] náḥamd73 Aḷḷāh yā wlēdī́... naḥmid Aḷḷāh... 31.) alḥuṛmah hēḏiy kānat zamān alliy bәygūluw lēhiy Šēxah bәtṭill ʿilayy... w ana wiǧʿānih, ḥuṛmah zēnih... (ʿĪ) ṣāḥbitkiy hēḏiy... (A) w Aḷḷāhiy ṣāḥǝbtī́ ḥuṛmah ṭayybih lākan aṣṢaṛāyʿih ʿamiriy fālhuṃ74. 32.) (ʿĪ) bass mā saṛaḥtiy maḥḥa intiy w iyyāha w intiy ṣġāyyrih... (A) lā! lillih75 ḥuṛmah hī. hī ǧat ḥuṛmah. 33.) ǧāybih ṯalāṯ әwlād... ʿAwdih... әw Han... ʿAwdih, әw ʿAwwād әw Ǧizzāʿ... (ʿĪ) әw Ǧizzāʿ...34.) (A) әw minnih ṭallagawhiy w axaḏhiy wāḥid ismih Rāḏ̣iy Aḷḷāh yaṛḥamih... әmn aṣṢaṛāyʿih... ǧābat әwlēd ismih Ramaḏ̣ān kibīr halḥīn... (ʿĪ) ạ̄, kibīr halḥīn Ramaḏ̣ān... 35.) (A) w Aḷḷah mā-ʿarif ʿind annās bәygūluw... (ʿĪ) mmm... (A) baʿadēn... (ʿĪ) әw minnih bәtṭill, kibrit halḥīn mā bәtṭill ʿalēkiy w intiy kibirtiy... 36.) (A) lā mā kabrit76, ǧawwazawhiy biʿīd... (ʿĪ) ǧawwazawha biʿīd... (A) w Aḷḷah w әʿwēlāthiy yā ṛayt yúkәbṛuw w әyṭilluw ʿalēna. 37.) (ʿĪ) wēn? әtlāwah Ṣadәr walla tlāwah Ḏaháb? (A) aḅūhiy fi Ṣadәr... (ʿĪ) aḅūha fi Ṣadәr... (A) ismih Swēlim әmn                                                                                                                         73 náḥamd Aḷḷāh: the imperfect of the verb ḥamad, yaḥmid “praise” receives an extra vowel a following non-final ḥ preceded by a (in accordance with the gahawah-syndrome). This is an indication that the gahawah-syndrome historically predates the rule for vowel harmony, which specifies that vowels in the verbal prefixes harmonize with the vowel in the stem of the verb (here i in ḥmid). Hence the imperfect form is neither yaḥamad (not possible since a in the final syllable would not have been elided) nor yiḥmid, but yaḥamid. The i in the final syllable in naḥamid Aḷḷāh is dropped from the open syllable in sandhi à náḥamd Aḷḷāh. Other similar examples are yaḥáriṯ “he ploughs”, yaʿárif “he knows”, yaxázin “he stores”. 74 ʿamriy fālhuṃ: was explained to me as a mild form of curse, something like “may their future be troubled”. 75 lillih: an emphatic negation typically used in women’s speech: “no way!”. 76 kabrit: the underlying a in kibir (|kabir|) reappears in closed syllable, see De Jong 2011: 250. Notice the difference between Tuṛbāniy and Tīhiy also in sentence 41 below.

78

Rudolf de Jong

adDrūz... (ʿĪ) әSwēlim әmn adDrūz... 38.) (A) әw lih axūh ismih Ḥamd... bass wāḥid kibīr, bass mā-ʿárif 77 ḥayyīn walla laʾ. 39.) әw lēš? alḥuṛmah ḥablíy 78 , әhniyyih yaḏ̣ḥakow ʿalēhiy, әw yiʿṭuw riǧǧāl ǝyṭallighiy... w aḅūhiy gāḷow ǧawwazhiy fi... ʿind Bīr alʿAbәd79... 40). (ʿĪ) halḥīn әtlāt Bīr alʿAbәd... (A) bәygūluw... awwil lay... halḥīn. halḥīn Taṛābīn māš šiy 80 , zamān bәyǧín alәḥṛayyim әgṛayybāt әw ġaṛṛabaw ʿind Zaṛānīg әhnúh. 41.) (ʿĪ) әw baʿadēn annās kubirit w alḥayāh xtálafat yā-ṃṃ Xiḏ̣әr... (A) yā Sā... (Sātir)! yā ʿĪd addinniy-xtaláfat... Aḷḷāh yā ṛabb yaṛḥamna b ṛaḥámtih w Aḷḷah-xtaláfat әw kabṛit addiniy w ʿamriy fālhuṃ, [әḏnūbhuṃ] šēnih... Translation: The Accident 1.) (ʿĪ) This accident, how (was it)? How dit it happen to you, oh Mḥammad? (M) This, I was locked up (in prison). I came out in ninety six, and then what? I went home to the north there. 2.) By God, and I stayed… so (then) I got a car… for twenty four… for thirty four thousand. This was happening in ninety six. 3.) And I brought the car going home… and the car (then) maybe what? A hundred meters from the house… I then said to the man, you get what’s coming to you (lit. your destiny/lot takes you). 4.) And the man and I, as we came driving, when we were driving the car, he said “I want to drive your car”, (so) I got out and got back into the car. 5.) We were three persons. I got in where? The outside (i.e. other) door… the side of the passenger. By God and I got in… when we approached the house… there was a patch of deep sand… 6.) I used to drive when I was a young boy, in the old days… I said “Do you know how to drive?” 7.) He said “I’m used to driving through deep sand that needs driving ability”. By God, when he went into the deep sand… 8.) From third (gear) - that is, it wasn’t going in the right gear – he shifted back to second. Because of that, the car took a turn. 9.) (ʿĪ) He lost control. (M) Yes, he like really lost control… When he lost control… the wheels came outside of the track… the front wheels… 10.) And immediately there was the explosion. The man died on the spot, and I… My two legs, when you touch them (lit. your hand on them) from here. And next to my eye here (lit. this), and (this) my hand (i.e. were hit                                                                                                                         77 mā-ʿárif: mā + aʿárif, see also preceding fn. 78 ḥablíy: the fem. sg. form of colours and physical defects has raised and stressed –íy (< *-āʾ) in neutral environments, see De Jong 2011: 235. 79 Bīr alʿAbd: a town in the north of Sinai, coordinates are appr. 31.00.54 N and 33.00.40 E., see Google Earth. 80 māš šiy: see Stewart 1990: 27 (text 8), line 26 (+ fn).

 

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sina

79

by shrapnel) 11.) By God, and I spent forty-five days in hospital. And after the forty five (days) 12.) (ʿĪ) In which hospital? (M) The hospital of Al-Arish… this general hospital. By God, and I stayed forty-five days and came out… 13.) And so after that what? When I got out I did my papers, I didn’t have my ID card with me, (so) then I did my ID, and I applied. 14.) He said “Shall we get you a wheel chair or artificial legs (lit. apparatuses)?” I said “No, let’s see the artificial legs…” If it had been a wheel chair, you would have found me now in a bad situation. 15.) By God, and I took the artificial legs… I threw them in the house for a month, until I got used to them afterwards. 16.) I am used to them now, I drive and I ride camels… that is, it was what? What is difficult for me is for instance running, I can’t run well or walk fast. As for the rest of the movements, they are okay… natural… 17.) (ʿĪ) This is from what was left behind of the mines. (M) The army of the days of sixty-seven. (ʿĪ) But you know that there are mines in this area. (M) I don’t know about them (precisely), I know whether there are mines or not… 18.) The way (i.e. it happened on the way). In old times the flood had brought them from up, with the earth in this throw it had brought them, the mines… So it had done what? It had thrown sand on top. 19.) Of course, because of this sand, the wind blows it up. What does it blow up? Then it comes near and it thins out the layer of land. The layer that is between the mine becomes… becomes very thin. 20.) It becomes when… when it is big (i.e. a thick layer of sand), the car doesn’t affect (lit. stir, arouse) it… Translation: The Mother of Xiḏ̣r 1.) (ʿĪ) Are you the mother of Xiḏ̣r? (A) Yes, by God. By God, the mother of Xiḏ̣r, may God keep my little son for me. May He let me see my little son. May He keep his little daughters for him. (Ḏ̣) […] (A) What, what? 2.) (ʿĪ) Oh, Amm Xiḏ̣r, where are you from originally? (A) I am originally from the ʿAwāmrih, the good people, but they died, may God have mercy on them… (ʿĪ) God have mercy on them. 3.) (A) May God have mercy on all, but… I am an orphan, an orphan. (ʿĪ) But those living in… (A) There’s three of us… Wardih, she’s the one walled in in Nixl81, and one with the Mzēnih… and we… 4.) And I was married a good Mzēniy man from, but he has gone to heaven, of the Ġṣayynāt from Ḏahab, good people. 5.) The father of this child died when he was young.82 And there I was trying to find an income (with the help) of God. And I have had a                                                                                                                         81 Nixl: the local pronunciation of the name Nakhl, a small town in central Sinai. 82 She is talking about Ḏ̣ayfaḷḷah, her son, who was present at the interview as well.

80

Rudolf de Jong

difficult life, and in the snow, I would move until the cold killed me now.83 6.) And I lived – God be praised – in these huts, as I ran up and I ran down and the (line of) partition stopped people there from (coming to) me.84 And I would carry bread from the wadi here, and the water here. 7.) And the firewood, but firewood was plentiful in the old days. And when we were unable (to move) because of the pain… 8.) If it were not for the child, my little son, who serves me… (ʿĪ) But from the inhabitants of where are you? (A) By God, from the day I was created… (I have been) in this land, in alĠaṛábah and in Fēṛān… in alĠaṛábah and in Fēṛān, and I (here)… 9.) from before the war… from before Ǧamāl died… (ʿĪ) Ǧamāl was the President of Egypt in the old days. (A) May God have mercy on him, Ǧamāl ʿAbd inNāṣir. (ʿĪ) Yes… 10.) (A) I was in Fēṛān, and I came to aṭṬaṛfa, and now… and after two years the war came. (ʿĪ) After two years after the war… (A) […] After two years…. 11.) (ʿĪ) Was that the war of sixty seven? (A) I was breast feeding Hlālih85, and I got Sālim in Fēṛān. And after three years the old man Sillām married me (lit. took me). 12.) And when I got Xiḏ̣r, I was still breast-feeding, the war of Israel against us started… (in) the year sixty seven, here in aṭṬaṛfa. (ʿĪ) The war of sixty seven… 13.) (A) And I was in this land all along… I never left it… I had an uncle who used to come to me, God have mercy on him, when he died, here I was… in my place… and he had children, and the people, the good children died… 14.) (ʿĪ) And you went to live… (A) Only I lived here… (ʿĪ) And how was Fēṛān in the days that you came here? (A) By God, I did not come there long ago. I had Xiḏ̣r and Ḏ̣ayfaḷḷah when I had not (yet) come there. I had them in this land. 15.) (ʿĪ) In this (lit. the) land (here), and you did not go down to the wadi? (A) No, no, no, no, no. I had Xiḏ̣r here in aṭṬaṛfa, and Ḏ̣ayfaḷḷah behind that tree trunk. 16.) And here I am in this land… with the Ǧbāliyyih, I have not passed… (ʿĪ) Yes… (A) The Ǧbāliyyih are good people, they don’t like problems… 17.) (ʿĪ) Okay, tell me… I want you to tell me how you make samn šīḥiy… (A) samn šīḥiy is good. (ʿĪ) Yes… (A) Oh Lord, may He grant us small cattle… (ʿĪ) Oh Lord… But how do they make it? 18.) They make it, they churn it… or do you want (it told from) the churning, or                                                                                                                         83 The recording was made in January in the al-Axaḏ̣ar area (coordinates of ʿAyn alAxaḏ̣ar are appr.28.49.00 N and 33.53.55 E, see Google Earth) and temperatures during the night could be well below zero in this location at approximately 1300 metres altitude. 84 Sinai was occupied by Israel from 1967 until 1982 and was returned to Egypt in stages from 1979-1982 as part of the Camp David Accords. 85 Hlālih is her daughter.

 

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sina

81

the mi…(ʿĪ) From the moment they have milked it… (A) When they have milked it, they clarify it… 19.) (ʿĪ) Yes… (A) And they churn it, and they make it into butter… (R) In the goat skin… (A) And they put it on the fire… yes, in the goat skin… and the make it into butter. 20.) When there is a lot of butter, they melt it… on the fire, and they add salt and curcumin, they filter it, and when it has cooled off they add… 20.) (ʿĪ) Do they put ziragrag on it? (A) What? (ʿĪ) Do they put ziragrag on it? (A) No, they put yellow curcumin on it… (ʿĪ) Yellow curcumin… 21.) (A) Yes, in the old days, when you were breast-fed… we had goats, (but) no more. […] Yes, yellow curcumin. They boil it with the salt in it…22.) and after that what? They carefully pour it (through a funnel) when it has cooled down (after a long time) in a clean receptacle… (ʿĪ) Of course it should be a clean receptacle. 23.) Yes, this is making curcumin, this is samn šīḥiy. (ʿĪ) This is samn… (A) This samn, we don’t eat (lit. taste) it, poor me, because of my chest, I don’t eat samn or (olive) oil. 24.) Because of the snow and cold that I saw in the old days… Poor me, the life of the children was in my hands. (ʿĪ) The life of what? (A) The life of my children was in my hands, the girl… (ʿĪ) Yes… 25.) And Xiḏ̣r and Ḏ̣ayfaḷḷah here in huts, I raised them with my hands. I used to go find firewood with people and bring it (home)… A life – God be praised - (ʿĪ) This firewood, what kind… (A) Firewood, nobody can find it for the fire, the cold has killed us… 26.) (ʿĪ) But in this land, what is the type of firewood? (A) This is ʿaǧram86. My little daughter went (out) yesterday, she selected it for me because I was in pain. 27.) (There was) ʿaǧram in the land, but it is not found (anymore), but they used it all. People (now) have paraffin, but we can’t afford it, we can’t get paraffin… (ʿĪ) draught, draught… the draught has slaughtered the land. 28.) (A) By God, it has cut us off from the land, oh ʿĪd. The cold even kills us at night. In the old days we had bʿayṯaṛān87. (ʿĪ) Yes…. (A) And retem wood, and its leaves, and wormwood, and ʿaǧram, and every tree/bush… 29.) (ʿĪ) God’s blessings… (A) In those mountains I would graze small cattle when I was strong, (but) the trees/bushes have disappeared and the small cattle and… 30.) (A) For ten years I have not gone down to the wadi. I have not gone down this wadi for fifteen years. Here I am, oh ʿĪd, I have been ill for ten days [coughs], in pain… [clears her throat] Let us praise God, oh little son… Let us praise God… 31.) This woman that they used to address as                                                                                                                         86 Probably Anabasis setifera. 87 The tubers (guʿṛah) of the Judean wormwood (Artemisia judaica) are used as firewood.

82

Rudolf de Jong

‘sheikha’ used to come and see me in the old days… when I was ill, a good woman… (ʿĪ) This was your friend… (A) By God (she was) my friend, a good woman, but may the future of the Ṣaṛāyʿih be troubled.88 32.) (ʿĪ) But you did not graze small cattle together with her when you were young… (A) No, not at all. She was a woman. She came as a woman.89 33.) She had three sons... ʿAwdih... and what’s his name... ʿAwdih, ʿAwwād and Ǧizzāʿ... (ʿĪ) And Ǧizzāʿ… 34.) (A) And after that they divorced her and somebody named Rāḏ̣iy married (lit. took) her - may God have mercy on him… - of the Ṣaṛāyʿih… she had a little boy named Ramaḏ̣ān… he is grown up now…(ʿĪ) Yes, he is not grown up, Ramaḏ̣ān… 35.) (A) By God, I don’t know with people, they say… (ʿĪ) mmm… (A) After that… (ʿĪ) And then she used to come and see you, she has now grown old and she not come to see you now that you have grown old… 36.) (A) No, she has not grown old. They married her off (to a man) far away. (ʿĪ) They married her off (to a man) far away. (A) By God, and may her little children, I wish, grow up and come and see us. 37.) (ʿĪ) Where (are they)? In the direction of Ṣadr or in the direction of Ḏahab? (A) Her father is in Ṣadr… (ʿĪ) Her father is in Ṣadr…(A) His name is Swēlim of the Drūz90. (ʿĪ) Swēlim of the Drūz… 38.) (A) And he has his brother named Ḥamd… But he is old, but I don’t know whether they are alive or not. 39.) And why? The woman is simpleminded, here they laugh at her, and they let her marry a man who divorces her… And they say her father gave her in marriage in… near Bīr alʿAbd… 40.) (ʿĪ) Now (she lives) in the direction of Bīr alʿAbd… (A) They say… at first… now. Now, there are no more Taṛābīn (to help me). In old times the women living near would come, and they went north near Zaṛānīg over there. 41.) (ʿĪ) And after that people grew old and the life has changed, oh mother of Xiḏ̣r… (A) Oh Lo(rd)! Oh ʿĪd, the world has changed… May God, oh Lord, have mercy on us with his mercy, by God, it has changed and the world has grown old, and may their future be troubled, [their sins/crimes] are bad…

                                                                                                                        88 Ṣaṛāyʿih: a clan of the Taṛābīn. 89 The woman was no longer a young girl, and therefore did not go out to graze small cattle. 90 The Drūz are also a clan of the Taṛābīn.

 

Texts in the Bedouin dialects of the Awlād Saʿīd and the Tayāha of Sina

83

Bibliography Aṭ-Ṭayyib, Muḥammad Sulaymān. 1993 (1st edition). Mawsūʿat alqabāʾil al-ʿarabīya, buḥūṯ maydānīya wa-tārīxīya. Cairo: Dār al-Fikr al-ʿArabī. __. 1997. (2nd edition). Mawsūʿat al-qabāʾil al-ʿarabīya, buḥūṯ maydānīya wa-tārīxīya. Cairo: Dār al-Fikr al-ʿArabī. Bailey, Clinton. 1985. “Dating the Arrival of the Bedouin Tribes in Sinai and the Negev”, Journal of the Economic and Social History of the Orient 28 (no.1): 20–49. __. 1991. Bedouin Poetry from Sinai and the Negev, Mirror of a Culture. Oxford: Clarendon Press. __. 2009. Bedouin Law from Sinai and the Negev, Justice without Government. New Haven and London: Yale University Press. Blanc, Haim. 1970. “The Arabic Dialect of the Negev Bedouins”, Israel Academy of Sciences and Humanities, Proceedings, Vol. 4: 112–150 (reprinted in Stewart 1990). EALL. 2006, 2007, 2008, 2009 (four volumes + index vol.) Encyclopedia of Arabic Language and Linguistics. Kees Versteegh (gen. ed.), Mushira Eid, Alaa Elgibali, Manfred Woidich, Andrzej Zaborski (ass. eds). Leiden – Boston: Brill. EALL Online. Encyclopedia of Arabic Language and Linguistics. Lutz Edzard and Rudolf de Jong (gen. eds.), Ramzi Baalbaki, James Dickins, Mushira Eid, Pierre Larcher, Janet Watson (ass. eds.). Leiden: Brill. (Online continuation of EALL, includes all lemmata of (printed) EALL). Henkin, Roni. 2010. Negev Arabic, Dialectal, Sociolinguistic and Stylistic Variation. Wiesbaden: Harrassowitz Verlag. Hopkins, Simon. 1990. “The word ġayr”, in Stewart 1990: 297–300. Jong, Rudolf E. de. 2000. A Grammar of the Bedouin Dialects of the Northern Sinai Littoral, Bridging the Linguistic Gap between the Eastern and Western Arab World. Leiden, Boston, Köln: Brill. __. 2011. A Grammar of the Bedouin Dialects of Central and Southern Sinai. Leiden, Boston: Brill. __. 2013. “Texts in Sinai Bedouin Dialects”. In: Ingham of Arabia, A collection of articles presented as a tribute to the career of Bruce Ingham. Clive Holes and Rudolf de Jong (eds.), 119–149. Leiden, Boston: Brill. Palva, Heikki. 1991. “Is there a North West Arabian dialect group?”. In: Festgabe für Hans-Rudolf Singer, zum 65. Geburtstag am 6. April 1990, überreicht von seinen Freunden und Kollegen I: 151–166. Frankfurt am Main: P. Lang. __. 2008. “Northwest Arabian Arabic”, Encyclopedia of Arabic Language and Linguistics. 400–408. Leiden, Boston: Brill. Procházka, Stephan. 2011. “Color terms”. In: EALL Online. Stewart, Frank Henderson. 1990. Texts in Sinai Bedouin Law, Part 2. The Texts in Arabic. Wiesbaden: Harrassowitz. Wehr, Hans. 1980. A Dictionary of Modern Written Arabic, third printing (reprinted in Beirut, by special arrangement with Otto Harrassowitz). Beirut: Librairie du Liban.

Ein Begleitbrief von 904 H zu Erlassen aus dem mamlūkischen Ägypten Werner Diem, Universität zu Köln

Das Dokument P.Vind. A.Ch. 36.580, das hier in Edition und Übersetzung vorgelegt wird, ist ein Begleitbrief, den der Schreiber, ein Beamter, zusammen mit Erlassen und dem dazugehörigen Brief eines bestimmten Emirs an einen Kollegen in der für den Vorgang zuständigen Behörde sendet. Das Dokument stammt gemäß seiner Datierung auf den 24. Ḏū l-Qaʿdah 904 H aus den letzten Jahrzehnten mamlūkischer Herrschaft in Ägypten. Der eben erwähnte Emir und ein weiterer hochgestellter Emir, den der Schreiber in seinem Schreiben erwähnt, tragen beide die Amtsnisbah aẓ-Ẓāhirī, die sich auf den mamlūkischen Herrscher al-Malik aẓ-Ẓāhir Qānṣūḥ bezieht. Dieser Herrscher trat am 17. Rabīʿ al-ʾawwal 904 die Herrschaft an (Ibn Iyās: Waqāʾiʿ III S. 405), und er beendete seine Herrschaft durch Flucht am 29. Ḏū l-Qaʿdah 905 (Ibn Iyās: Waqāʾiʿ III S. 436). Das Dokument ist also gut acht Monate nach Amtsantritt von al-Malik aẓ-Ẓāhir Qānṣūḥ geschrieben worden. Das Datum, 24. Ḏū l-Qaʿdah 904 H, entspricht dem 3. Juli 1499 n. Chr. Zur Formalbeschreibung des Stücks ist anzumerken, daß es sich um gelblich-braunes Papier und braune Tinte handelt. Das Papier ist bis auf Stücke in der linken oberen Ecke und im linken unteren Bereich sowie in Z. 4 bei geringem Textverlust vollständig. Es ist oben ein schmaler und rechts ein breiter Rand gelassen. Verso ist leer. Das Blatt war mehrmals quer gefaltet. Die Größe beträgt 21,6 x 12,7 cm.

Ein Begleitbrief von 904 H zu Erlassen aus dem mamlūkischen Ägypten

85

86

Werner Diem

Was die Interpretation im einzelnen betrifft, so muß vorausgeschickt werden, daß das Schreiben nicht nur wegen extremer Kursivität wie auch individueller Züge des Duktus große Probleme aufwirft, sondern auch deshalb, weil jede Selbstbezeichnung des Schreibers und jede namentliche Erwähnung des Adressaten fehlen. Der Adressat wird nur an zwei Stellen mit dem titelartigen Ausdruck al-ḫidmah al-karīmah (Z. 3) und ḫidmatuhu l-karīmah (Z. 6), angesprochen, was „der edle Dienst“ bzw. „sein edler Dienst“ im Sinne von „das edle Amt“ bzw. „sein edles Amt“ bedeutet. Entgegen dieser formal femininen Anrede wird der Adressat ansonsten ad sensum in der 3. Mask. Sing. angesprochen. Auf sich selbst referiert der Schreiber, mit einer Ausnahme gegen Ende des Schreibens, wo er in einer Art Captatio benevolentiae etwas persönlicher wird, überhaupt nicht. An dieser letzteren Stelle spricht er von sich als „uns (Dativ)“ (Z. 11), wobei der Plural als Bescheidenheitsform aufzufassen ist. Phraseologische Anklänge an bereits edierte mamlūkische amtliche Schreiben bestehen, aber das Schreiben bildet in phraseologischer Hinsicht doch bis auf weiteres weitgehend ein Exemplar sui generis. Um den Brief zu verstehen, muß erst etwas zum Hintergrund gesagt werden. Bekanntlich bezogen die mamlūkischen Militärs Ägyptens anstatt eines Solds Einkünfte aus Lehen (ʾiqṭāʿ), die aus dem festgesetzten Steuerwert (ʿibrah) ägyptischer Dörfer bestanden. Bei dem vorliegenden Vorgang geht es um Steuereinkünfte aus drei Dörfern namens Duwaynah, Mūšah und Sidfah (Z. 4). Alle drei Dörfer sind in der oberägyptischen Provinz al-Aṣyūṭ gelegen. Die im Dokument an zweiter und dritter Stelle genannten Dörfer, Mūšah und Sidfah, sind in mamlūkischen Lehensregistern namentlich erwähnt (Halm: Lehensregister I S. 93 und S. 97). Das im Schreiben an erster Stelle genannte Duwaynah, das in mamlūkischen Lehensregistern nicht namentlich erwähnt wird, ist nach Ramzī: Qāmūs II 4, S. 18 mit dem Ort Ṭūḫ Bakrīmah (Halm: Lehensregister I S. 98) identisch. Nach Ramzī ist der Ortsname Ṭūḫ Bakrīmah im Steuerkataster von 933 H, der unter osmanischer Herrschaft erstellt wurde, durch den Ortsnamen Duwaynah ersetzt, doch bietet Ramzī für den Namen Duwaynah bereits aus dem 9. Jahrhundert H Nachweise. Das vorliegende Dokument zeigt, daß die offizielle Umbenennung des Ortes Ṭūḫ Bakrīmah zu Duwaynah bzw. die steuerliche Einbeziehung von Ṭūḫ Bakrīmah unter Duwaynah bereits vor 933 H vorgenommen worden sein muß. Auf Karte 8 in Halm: Lehensregister I liegen die drei Orte Mūšah, Duwayna /Ṭūḫ Bakrīmah und Sidfah auf einer schrägen Linie, die man westlich des Nils von der Provinzhauptstadt al-Aṣyūṭ im Nordwesten nach Sidfah im Südosten ziehen könnte.

Ein Begleitbrief von 904 H zu Erlassen aus dem mamlūkischen Ägypten

87

Für alle drei Dörfer geben die Lehensregister für das Ende des 9. Jahrhunderts H, also für einen nur wenig vor dem Vorgang liegenden Zeitraum, an, daß ihre Steuereinkünfte an mehrere Lehensträger sowie Stiftungen flossen. Bei einer solchen Aufsplitterung der Steuereinkünfte in Bruchteile war natürlich nicht daran zu denken, als begünstigter Lehensträger selbst die Erhebung der Steuern zu organisieren. Dies war im allgemeinen selbst dann schwierig, wenn dem Lehensträger ein ganzes Dorf zustand, weil ein einzelnes Dorf den Unterhalt einer eigenen Verwaltung nicht trug (Halm: Lehensregister I S. 17). Wie in solchen Fällen von Bruchteilseinkünften verfahren wurde, ist nicht im einzelnen bekannt. Eine Lösung bestand vermutlich darin, daß ein Emir, der ein umfangreiches Lehen in der Nähe solcher Dörfer besaß, diese gegen Entgelt mitverwaltete und die Beträge mit den Begünstigsten abrechnete. Eine andere Lösung könnte darin bestanden haben, daß ein Steuerpächter die Dörfer übernahm und den Begünstigten die auf sie entfallenden Beträge auszahlte. Eben diese zweite Möglichkeit wird durch das Begleitschreiben konkretisiert. Die Erlasse, von denen in dem Schreiben die Rede ist, hatte nämlich offenbar der Steuerpächter in Sachen ausstehender restlicher Steuern erwirkt. Da er im Schreiben bereits in der zweiten Zeile als „der Erwähnte“ eingeführt wird, muß man annehmen, daß der Adressat mit dem Vorgang vertraut war; vielleicht hatte es eine vorhergehende Beschwerde gegeben. Aus dem Schreiben wird deutlich, daß der Steuerpächter in Kairo ansässig war oder dort wegen seiner Beschwerde weilte, und daß der Schreiber seinen Begleitbrief aus Kairo an den Adressaten richtete, während der Adressat ein leitender Beamter in der Provinz al-Aṣyūṭ gewesen sein muß. Der Steuerpächter verfügte aber offenbar seinerseits über keine eigene Organisation zur Eintreibung der Steuern, sondern dies oblag, wie aus dem Schreiben hervorgeht, der Finanzbehörde. Insofern „schuldete“ die Finanzbehörde, wie es im Schreiben heißt (Z. 7), dem Steuerpächter die restlichen Steuern der drei Dörfer. Allerdings überrascht es, daß die zuständige Finanzbehörde im Brief als „Ministerium für [Ober]ägypten“ bezeichnet wird (Z. 4), da man annehmen sollte, daß die Eintreibung der Steuern bei der Finanzverwaltung der Provinz al-Aṣyūṭ hätte liegen müssen. Dieses Problem löst sich aber, wenn man berücksichtigt, daß al-Aṣyūṭ der Sitz des Gouverneurs von Oberägypten war und über keinen eigenen Provinzgouverneur verfügte (Halm: Lehensregister I S. 61). Das Ministerium für Oberägypten wird deshalb zugleich für die Steuerverwaltung der Provinz al-Aṣyūṭ zuständig gewesen sein, und in eben dieser Behörde muß der Adressat des Schreibens an maßgeblicher Stelle tätig gewesen sein.

88

Werner Diem

Die Erlasse, von denen im Schreiben die Rede ist, hatte der Steuerpächter, wie erwähnt, in Kairo erwirkt. In diesen Erlassen wird angeordnet, daß ein in Kairo in unbekannter Funktion (der Text scheint hier lückenhaft zu sein) tätiger Emir namens Titel Sayf ad-Dīn Arūk sich des Steuerpächters annimmt (Z. 1f.), und von diesem Emir scheint auch ein Brief zu stammen, der zusammen mit den Erlassen beim Schreiber eingetroffen war (Z. 2). Da der Schreiber einleitend zweimal davon spricht, daß „die“ Erlasse angekommen seien (Z. 1, 3), die Erlasse dem Adressaten aber zu dem Zeitpunkt, da der Schreiber seinen Brief verfaßte, nicht bekannt sein konnten, ist anzunehmen, daß der Schreiber die Erlasse nebst dem erwähnten Brief des Emirs Sayf ad-Dīn Arūk dem Adressaten zusammen mit seinem eigenen Begleitschreiben übersandte. Wenn der Adressat die Erlasse zusammen mit dem Begleitschreiben des Absenders erhielt, waren sie für ihn in der Tat „die“ Erlasse, weil sie ihm konkret vorlagen. Da der Schreiber dem Adressaten, die Erlasse resümierend, dringend nahelegt, sich des Steuerpächters bis zum Abschluß der Angelegenheit anzunehmen, hat der Steuerpächter die Unterlagen wohl selbst überbracht. Abgesehen von dem Emir Sayf ad-Dīn Arūk, der seiner Titulatur gemäß einen mittleren Rang innehatte, wird in dem Schreiben noch ein weiterer, sehr viel höher stehender Emir erwähnt, welcher der Befehlshaber der königlichen Leibgarde war. Im Dokument hat er die Amtsbezeichnung ʾamīr an-nawbah assulṭānīyah (Z. 6). Es scheint, daß der Steuerpächter die Protektion dieses Emirs genoß, weil der Schreiber erwähnt, daß die Erlasse der Signatur dieses Emirs teilhaftig geworden seien (Z. 5), sei es, daß sie von ihm stammten oder daß er sie auf dem Dienstweg abgezeichnet hatte, und weil der Schreiber den Adressaten außerdem im weiteren Verlauf des Schreibens auffordert, den Steuerpächter entsprechend seiner Verbindung zur königlichen Leibgarde zu behandeln (Z. 7) und ihm jede Hilfe zu leisten, bis er nach Abschluß der Angelegenheit zur Leibgarde zurückkehren werde (Z. 10). Man muß sich hier die Frage stellen, welcher Art die Verbindung zwischen einem hochgestellten Emir, wie es der Befehlshaber der Leibgarde des Herrschers war, und einem Steuerpächter gewesen sein könnte, die den Emir veranlaßte, sich des Steuerpächters anzunehmen. Die Antwort könnte in dem bereits erwähnten Umstand liegen, daß die Steuereinkünfte aus den drei Dörfern auf verschiedene Emire und Stiftungen aufgeteilt waren. Es wäre denkbar, daß einer oder mehrere jener Emire der herrscherlichen Leibgarde angehörten, so daß dem Kommandanten der Leibgarde daran gelegen sein mußte, daß der Steuerpächter seine Einnahmen erhielt, weil er sonst seinen Zahlungsverpflichtungen nicht nachkommen konnte.

Ein Begleitbrief von 904 H zu Erlassen aus dem mamlūkischen Ägypten

89

Beim Amtsantritt al-Malik aẓ-Ẓāhir Qānṣūḥs (17. Rabīʿ al-awwal 904) war der Befehlshaber der königlichen Leibgarde Qurqmās min Arikmās min Walīy ad-Dīn. Dieser war im Šawwāl 903 H von al-Malik aẓ-Ẓāhir Qānṣūḥs Vorgänger, al-Malik an-Nāṣir Muḥammad, in das Amt des raʾs an-nawbah al-kubrā eingesetzt worden (Ibn Iyās: Badāʾiʿ III S. 391), und er behielt dieses Amt unter al-Malik aẓ-Ẓāhir Qānṣūḥ bis zum Šaʿbān 905 H, als er zum Statthalter von Aleppo ernannt wurde (Ibn Iyās: Badāʾiʿ III S. 431). Die Amtsbezeichnung Qurqmās min Arikmās min Walīy ad-Dīns variiert bei Ibn Iyās: ar-raʾs annawbah1 al-kubrā (III S.   391, 431), ar-raʾs nawbah al-kubrā (III S. 431) und raʾs nawbat an-nuwab (III S.   396, 408, 414, 416). Alle diese Varianten, von denen alQalqašandī nur den Terminus raʾs nawbat an-nuwab anführt (Ṣubḥ V S. 428), bezeichnen den Befehlshaber der königlichen Leibgarde. Im vorliegenden Dokument wird hierfür der Ausdruck ʾamīr an-nawbah as-sulṭānīyah gebraucht. Obwohl die Identifizierung des im Dokument genannten Befehlshabers der Leibgarde mit Hilfe von Ibn Iyās’ Chronik möglich ist, läßt sich kein zwingender sprachlicher Konnex zwischen den Angaben des Dokuments und der Chronik herstellen, da einerseits Qurqmās min Arikmās min Walīy ad-Dīn bzw. Qurqmās min Walīy ad-Dīn, wie er meistens verkürzend genannt wird, in Ibn Iyās’ Chronik stets nur mit seinem Namen ohne Ehrentitel und andererseits der mit ihm vermutlich identische Emir des Dokuments nur mit seinen Ehrentiteln ohne Namen erwähnt wird. So lassen sich zwar die beiden Angaben kombinieren, bestätigen sich aber nicht gegenseitig. Immerhin kann als Stützung der Identifizierung noch der Titel al-maqarr, den der Emir im Dokument trägt, angeführt werden, da dieser Titel nach al-Qalqašandī den höchsten Würdenträgern im Staat vorbehalten war (Ṣubḥ VI S. 105), zu denen, auch wenn al-Qalqašandī dies nicht explizit erwähnt, sicherlich der Befehlshaber der königlichen Leibgarde gehörte, zumal der Titel, wie Bāšā (Alqāb S. 489ff.) ausführt, im Laufe der Zeit etwas inflationiert wurde. Konkret lautet der Titel im Dokument al-maqarr al-ʿālī, was gemäß al-Qalqašandī (Ṣubḥ V S. 464; VI S. 97, 105) die rangniedrigste oder eine der rangniedrigsten der Kombinationen mit dem Grundwort almaqarr ist. In einer Inschrift trägt der Befehlshaber der königlichen Leibgarde hingegen den höheren Titel al-maqarr al-ʾašraf, aber mit einer ganz ähnlichen Eulogie wie im Dokument: ‫صاره‬#‫ز اهلل أن‬#‫نصورى أع‬#‫ة امللكى امل‬#‫وب‬#‫ة رأس ن‬#‫رك‬#‫زينى ب‬#‫رف ال‬#‫قر األش‬#‫امل‬

1

‫النوبة‬: Druck ‫ النوبية‬, was wohl ein Druckfehler ist.

90

Werner Diem

RCEA XVII Nr. 782004 (al-Karak, 782 H)2. Im vorliegenden Dokument lautet die Eulogie ʾaʿazza llāhu nuṣratahu. Nach dem Appell des Schreibers an den Adressaten, dem Betreffenden, also wohl dem Steuerpächter, bei der Eintreibung der Rückstände jede erdenkliche Hilfe zuteil werden zu lassen, wie man es von ihm gewohnt sei (Z. 6–11), folgt die Aufforderung, ihn seinerseits mit Erlassen zu beehren (Z. 11). Eine Schlußeulogie (Z. 11f.), das Datum, religiöse Formeln und ein Schlußzeichen (Z. 13– 17), wie sie aus mamlūkischen Erlassen bekannt sind, beschließen den Brief. Warum in der Angelegenheit mehrere Erlasse ergangen waren, bleibt unklar. Deutlich wird aber jedenfalls, daß hier ein Vorgang gemäß einem genauen Prozedere von Behörde an Behörde weitergegeben wurde, bis er die zuständige Behörde erreichte. Obwohl es sich somit um einen festgelegten Dienstweg handelte, war es, um den Vorgang letztlich konkret durchzuführen oder um ihn zu beschleunigen, dennoch nötig, daß sich auf einer mehr kollegialen Ebene ein Beamter bei seinem zuständigen Kollegen dafür verwendete. Das Dokument ist, wie eingangs erwähnt, in einem sehr kursiven, typisch spätmamlūkischen Duktus geschrieben. Zu den Besonderheiten des Duktus gehören die teilweise minimale Höhe von Hasten (z.B. ‫لس‬# # # # # # ‫ املج‬Z. 1), die Überlänge von initialen Häkchen, die häufige Vereinfachung von rechtsverbundenem wāw zu einer rāʾ-ähnlichen Form ohne Kopf, die Überlänge von finalem yāʾ und nūn und die Plazierung von initialem überlangem sīn über die vorhergehenden Buchstaben (‫يم‬# ‫راس‬# ‫ امل‬Z. 1, 3, 6, 9; ‫يم‬# ‫مراس‬# ‫ ب‬Z. 11; ‫تخراج‬# ‫ اس‬Z. 7; ‫يحيط‬# ‫ س‬Z. 8; ‫مول‬# ‫ وش‬Z. 9; ‫تمرار‬# # ‫ واس‬Z. 10), ferner die winzige Größe von ‫ك‬# # ‫( ذل‬Z. 1, 9, 10) und die extrem vereinfachte Form von ‫ مع‬Z. 11, bei der der Kopf des ʿayn praktisch entfallen ist.

2

Kalus übersetzt ‫نصورى‬# # # # # # ‫ة امللكى امل‬# # # # # # ‫وب‬# # # # # # ‫ رأس ن‬zutreffend mit „capitaine de la garde d’alMalik al-Manṣūr“. Konkret fasse ich ‫نصورى‬# # # # ‫ة امللكى امل‬# # # # ‫وب‬# # # # ‫ رأس ن‬als eine Nisbah auf, die aus raʾsu nawbati l-Maliki l-Manṣūri gebildet ist, indem die beiden letzten Wörter in Nisbah-Bildungen transformiert wurden; diese aus vier Wörtern bestehende Nisbah ist zur Nisbah az-Zaynī parallel. Somit ist ‫ورى‬#‫ص‬#‫ن‬# # # # # # ‫كى امل‬#‫ل‬#‫ة امل‬# # # # # # ‫وب‬# # # # # # ‫ رأس ن‬genau genommen als raʾsu nawbati l-Malakīyu l-Manṣūrīyu zu transkribieren. – Zum Emir Zayn ad-Dīn Barakah siehe Ibn Taġrī Birdī: Manhal Nr. 654.

‫‪91‬‬

‫‪Ein Begleitbrief von 904 H zu Erlassen aus dem mamlūkischen Ägypten‬‬

‫‪Text‬‬

‫‪ ١‬سبب ذلك وردت املراسيم الشريفة بتحمل املجلس العالى االميرى الكبير]ى‬

‫[‬

‫‪ ٢‬السيفى انوك فى طية املعزية الظاهرى ادام اهلل تعالى نعمته الحضار املذكور الى‬ ‫]الخدمة الكريمة وقد[‬ ‫‪ ٣‬وردت املراسيم الشريفة بكتابه مما اورد الى الخدمة الكريمة فى هذا العني وكتب‬ ‫علـ}ـى{ يه انصافه تتمة ]ما[‬ ‫‪ ٤‬متاخر ومن الحق على نواحى ضمانه من ديوان الوجه القـ]ـبلى[ تتمة دوينة وموشة‬ ‫وسدفة الى‬ ‫‪ ٥‬اخر تودية املبلغ وتعميله )؟( للخراجية مشمولة بخط املقر العالى االميرى الكبيرى‬ ‫املخدومى االمينى ]االوحـ[ـدى املبارزى‬ ‫‪ ٦‬مقعد الدولة امير النوبة السلطانية برهان اهلل الظاهرى اعز اهلل نصرته واكدت املراسيم‬ ‫الشريفة على خدمته الكريمة >فى ذلكتودجو< االحاطة‬ ‫‪ ٩‬بذلك والعمل بما برزت به املراسيم الشريفة فى ذلك وشمول االقصارات بالنظر السعيد‬ ‫‪ ١٠‬واالعانة الشاملة واستمرار نظره السعيد عليه الى حني نهاية ذلك وعوده الى النوبة‬ ‫الشريفة بنضر>ه< السعيد‬ ‫‪ ١١‬كما هو املعود من تفضالته مع التشريف بمراسيم تصدر لنا و}و{العمل بها واهلل تعالى‬ ‫‪ ١٢‬يديم على العالية مواد نعمه‬ ‫‪١٣‬‬

‫ان شا اهلل تعالى‬

‫بمنه وكرمه‬

‫كتب فى رابع عشرين ذى القعدة الحرام‬

‫‪١٤‬‬ ‫‪ ١٥‬والحمد هلل وحـ]ـده[‬

‫عام اربع وتسعماية‬ ‫وصلى اهلل على سيدنا مـ]ـحمد و[اله وسـ]ـلم تسليما[‬

‫‪١٦‬‬

‫وحسبنا اهلل وحده‬

‫‪١٧‬‬

‫‪Schlußzeichen‬‬

92

Werner Diem

Übersetzung 1

Der Grund für dieses (Schreiben) ist: Es sind die (beigefügten) edlen3 Erlasse angekommen, daß die hohe Exzellenz des groß[en, 2–3 Epitheta] Emirs

2

Sayf ad-Dīn Anūk im Muʿizzischen (Kairo)4 aẓ-Ẓāhirī – Gott, der Erhabene, schenke ihm weiterhin seine Gnade – es übernehmen wird, den Erwähnten5 hinbringen zu lassen zu [dem edlen Amt6 , und zwar]

3

sind die edlen Erlassen zusammen mit dessen7 Brief angekommen, was (hiermit) an sein edles Amt übersandt wird, bezüglich dieses Geldes8, wobei ihm9 (darin) vorgeschrieben wird, jenem10 gerechterweise den Restbetrag dessen zu verschaffen, was

4

noch aussteht und (was) ein (steuerlicher) Rechtsanspruch ist gegen die Distrikte seiner Steuerpacht vom Dīwān des [südlichen] Gebiets, (nämlich) den Restbetrag (der Steuern) von Duwaynah, Mūšah und Sidfah, bis

5

zur endgültigen Bezahlung und Verbuchung des Betrags für das (betreffende) grundsteuerpflichtige (Land), wobei ihnen11 zuteil geworden ist die Signatur der hohen Exzellenz des großen Emirs, des Herrn, des Zuverlässigen, des Ein[zigart]igen, Mubāriz ad-Dīn,

6

Maqʿad ad-Dawlah, des Befehlshabers der herrscherlichen Leibgarde, Burhān Allāh aẓ-Ẓāhirī – Gott, der Erhabene, gewähre ihm mächtige Hilfe! Die edlen Erlasse sind gegenüber seinem edlen Amt nachdrücklich ,

3 Das Epitheton weist auf Erlasse des Herrschers oder einer sehr hochgestellten Person, vielleicht des in Z. 5f. genannten Emirs der königlichen Leibgarde. 4 Es scheint nach dem Namen des Emirs die Bezeichnung eines Amtes ausgefallen zu sein, das er in Kairo innehatte. 5 Es handelt sich um den Steuerpächter. Der Schreiber bezieht sich auf die Erlasse, in denen jener erwähnt war. 6 Aus dem Kontext ergänzt. Hier und später Bezeichnung des Adressaten, eines Kollegen des Schreibers. 7 Sc. des vorgenannten Emirs Sayf ad-Dīn. 8 Daß es um die Eintreibung von Geld, und zwar Steuerrückständen, ging, war in den Erlassen erwähnt, deren Kenntnis der Schreiber voraussetzt. 9 Sc. dem Adressaten. 10 Sc. dem Steuerpächter. 11 Sc. den in Z. 1 und 3 erwähnten Erlassen.

Ein Begleitbrief von 904 H zu Erlassen aus dem mamlūkischen Ägypten

93

7 nämlich, gemäß der Verbindung des Erwähnten12 zur edlen Leibgarde, indem er dies zur Kenntnis nimmt, zu verfahren, das einzutreiben, was noch aussteht und was rechtlicher Anspruch gegen den erwähnten Dīwān 8 bis zur endgültigen lung und Verbuchung des (betreffenden) Betrags ist, und das (zu tun), was sein edles Wissen bei der Einsichtnahme in es13 zur Kenntnis nehmen wird, wobei gegenüber niemandem Nachsicht in Hinblick auf seine Frist (für die Bezahlung der Steuer) besteht. Das Er ist, dies zur Kenntnis zu 9 nehmen, gemäß dem zu verfahren, was die ergangenen edlen Erlasse diesbezüglich beinhalten, Unzulänglichkeiten seine glückliche Fürsorge zuteil werden zu lassen14, 10 umfassende Hilfe zu leisten und seine glückliche Fürsorge fortgesetzt auf ihm15 ruhen zu lassen bis zu dem Zeitpunkt, da dies beendet ist und er zurückkehrt zur edlen Leibgarde auf Grund glücklichen Fürsorge, 11 wie man es von seinen Freundlichkeiten gewohnt ist. Es möge Beehrung durch Erlasse (von seiner Seite) erfolgen, die an uns ergehen und denengemäß gehandelt werden wird. Gott, der Erhabene, 12 gewähre der hohen (Exzellenz) weiterhin die Elemente seiner Gnade, so Gott, der Erhabene, will, in seiner Gunst und Großmut! 13 Geschrieben am vierundzwanzigsten des sakrosankten (Monats) Ḏū lQaʿdah 14 im Jahre neunhundertundvier. 15 Lob sei Gott al[lein]! Gott segne unseren Herrn Mu[ḥammad und] seine [Sip]pe [und] spen[de ihnen Heil!] 16

Uns genügt Gott allein!

17

Schlußzeichen

12 Sc. des Steuerpächters. 13 Entweder fehlerhafter Bezug auf die Erlasse oder aber Bezug auf einen der Erlasse oder das Schreiben des Emirs Sayf ad-Dīn. 14 Sc. Unzulänglichkeiten etwa der Steuerbeamten abzustellen. 15 Sc. dem Steuerpächter.

94

Werner Diem

Kommentar 1 ‫ك‬# # # # # # ‫ب ذل‬#‫ب‬# # # # # # ‫ س‬: Verwandte Ausdrücke sind wa-sababu katbi hāḏihi l-ḫidmati AGBW 44r, 28f.; sababuhā ʾilayhim ʾiʿlāmuhum Goitein: „Miktab“ 5, 7ff.; as-sababu ʾaḏkuru laka ʾamra Sihāmin ABHeid I 47, 4. Der letzte Beleg zeigt dieselbe Asyndese des Prädikats wie die vorliegende Stelle. Daß der Schreiber sich auf das eigene Schreiben mit ḏālika bezieht, überrascht; man würde eher hāḏā erwarten. Aber die Lesung ‫ك‬# # # # # # ‫ ذل‬ist zwingend, da dieselbe Form des Wortes in Z. 9 und 10 begegnet, an der zweiten Stelle ohne den kāf-Querstrich. Eine Lesung als ‫تبه‬# # # # ‫ ك‬ist nicht möglich. 2 Auf ‫ السيفى‬folgt ein Graphem, das am ehesten als ‫وك‬# # # # # # ‫ ان‬Anūk interpretiert werden werden kann. Siehe zu diesem Namen Sauvaget: „Noms“ Nr. 33 und Ibn Taġrī Birdī: Manhal Nr. 550f. – ‫ة‬# #‫عزي‬# #‫ية امل‬# #‫فى ط‬: Dieselbe Form von langgestrecktem flachem ‫ فى‬findet sich in Z. 8 und 9. Ein Beleg für fī ṭayyati in der ursprünglichen Bedeutung des Eingefaltetseins ist wa-l-mudraǧatu fī ṭayyatihi „und das in diesen (Brief) eingefügte (Schreiben)“ DAAF S. 26, pu.; das synonyme fī ṭayyi begegnet in bereits übertragener Bedeutung in wa-fī ṭayyi l-qinyāni minhu HJES III 34, 22. Warum diese Ortsbestimmung in der Titulatur des Emirs steht, ist mir unklar. Vielleicht ist davor eine Amtsbezeichnung ausgefallen. al-Muʿizzīyah ist eine Bezeichnung Kairos; siehe AGBW S. 234. – ‫ور‬#‫ذك‬#‫ضار امل‬#‫الح‬: ḥḍr IV -hu „erscheinen lassen; bringen“ bezeichnet auch das Vorführen einer Person bei einer Behörde; zu einigen Stellen siehe ABB, Glossar s.v. ḥḍr. Hier ist dies aber nicht in detrimentalem Sinn gemeint, da es sich um den Steuerpächter, handelt, dem der Rest der ihm zustehenden Steuern verschafft werden soll. Es ist also die Überstellung einer Person auf amtlichem Weg von einer Behörde zu einer anderen gemeint, hier vielleicht im Sinne von Avisierung. 3 ‫ اورد‬: Dieselbe Ligatur mit dāl findet sich in ‫ ادام‬Z. 2, ‫ود‬#‫ع‬# # # # # # ‫ امل‬Z. 11 und ‫واد‬# # # # # # ‫م‬ Z.   12. wrd IV -hu bezeichnet auch speziell das Abliefern von Briefen. – Mit alḫidmah al-karīmah ist der Adressat gemeint. Normalerweise bezeichnet ḫidmah einen eigenen Dienst des Schreibers und gegenüber höhergestellten Adressaten speziell den eigenen Brief. Hier ist der Dienst des Adressaten gegenüber dem Herrscher gemeint, also das Amt des Adressaten. – fī hāḏā l-ʿayni: Mit ʿayn wird in amtlichem Kontext Geld, gewöhnlich in Goldwährung, bezeichnet, üblicherweise im Zusammenhang mit Steuern. Daß der Schreiber hier von „diesem Geld“ spricht, zeigt, daß er die Kenntnisnahme des Adressaten von den Erlassen und dem ihnen beigefügten Brief des Emirs voraussetzt. – ktb I -hu ʿalā in der Bedeutung „etwas gegen jemanden festsetzen; jdm. etwas vorschreiben“ ist

Ein Begleitbrief von 904 H zu Erlassen aus dem mamlūkischen Ägypten

95

zwar unbedenkliches Arabisch, für ein Dokument dieser Art aber ungewöhnlich. Die Verbesserung von ‫ على‬zu ‫ليه‬# # # # # #‫ ع‬durch Anhängung von ‫ه‬# # # # # #‫ ں‬ist vermutlich dadurch entstanden, daß der Schreiber zunächst vorgehabt hatte, ‫ه‬#‫ت‬# # # # # # ‫دم‬# # # # # # ‫لى خ‬#‫ع‬ ‫ة‬#‫م‬# # # # # # ‫ري‬#‫ك‬# # # # # # ‫ ال‬wie in Z. 6 zu schreiben. – nṣf IV -hu -hu hat hier die Bedeutung „jdm. etwas gemäß einem rechtlichen Anspruch geben“. – tatimmah, etwa „Ergänzung, Komplettierung“, ist hier der für die volle Bezahlung der Steuerschuld noch ausstehende Teil, also der Restbetrag. In ABH 18, 7 ist von tatimmat an-niṣf die Rede, womit derjenige Restbetrag gemeint ist, der zur Hälfte des Gesamtbetrags noch fehlt. Ähnlich lautet die Überschrift in der Steuerabrechnung APEL 277 ǧarīda-tun bi-mā baqiya li-tatimmati naǧmi Ḏī l-Ḥiǧǧati sana[ta „Blatt mit dem, was noch fehlt für die volle Rate des Monats Ḏū l-Ḥiǧǧah des Jahr[es X“. 4 Mit ḥaqq „Rechtsanspruch“ sind hier wie auch sonst in amtlichen Schreiben die Steuern gemeint. – Da die im folgenden genannten drei Dörfer zu Oberägypten gehören, muß das Wort nach ‫ه‬# # #‫وج‬# # #‫ ال‬zu ‫ الـ]ـقبلى‬ergänzt werden. Der Kopf des qāf scheint noch teilweise erhalten zu sein. – Das darauf folgende ‫تمة‬# # ‫ ت‬nimmt das vorhergehende ‫تمة‬#‫ ت‬auf. Gemeint ist die Bezahlung des Restbetrags der Steuern der drei Orte. – Zu den Dörfern Duwaynah, Mūšah und Sidfah siehe die Einleitung. 4–5 ‫ية‬#‫راج‬#‫عميله للخ‬#‫ة املبلغ وت‬#‫ودي‬#‫ر ت‬#‫ الى اخ‬: Von diesem Passus hat ‫ة‬#‫ودي‬#‫ ت‬ein punktiertes tāʾ und ist insofern eindeutig, auch wenn das dāl abweichend vom üblichen Duktus des Schreibens ohne Schleife angeschlossen ist. tawdiyah ist Infinitiv zur Vulgärform wdy II < ʾdy II und steht für taʾdiyah. Siehe zu wdy II AGBW S. 149, Kommentar zu 30r, 5, und ein weiterer Beleg ist waddīhi „gib es her!“ AABW 47r, 6. Derselbe Passus erscheint in Z. 8 in der Form ‫عميله‬# # # ‫ة املبلغ وت‬# # # ‫ودت‬# # # ‫ الى اخ‬, d.h. mit Auslassung eines Teils von ‫ة‬# # # # # # ‫ودي‬# # # # # # ‫ت‬, was ein Schreibfehler ist, und ohne ‫ية‬# # # ‫راج‬# # # ‫للخ‬. Auf ‫ املبلغ‬folgt an beiden Stellen unpunktiertes ‫عمله‬# # # ‫وں‬, dessen Pronomen sich auf ‫لغ‬#‫ب‬#‫ امل‬bezieht und dessen Sinn dem Kontext gemäß „und Verbuchung desselben“ sein muß. Allerdings ist in dieser Bedeutung nur der I. Stamm von ʿml üblich. Ich nehme dennoch an, daß hier der II. Stamm in derselben Bedeutung in kursiver Schreibung gemeint ist. Das Wort ‫ية‬# # # # # #‫راج‬# # # # # #‫ الخ‬erscheint gelegentlich als Verkürzung von ‫ية‬# # #‫راج‬# # #‫ االرض الخ‬als Bezeichnung von grundsteuerpflichtigem Land, so in APEL 261, 10. Es ist aber auch nicht ausgeschlossen, daß ‫ية‬#‫راج‬#‫الخ‬ hier für ‫ية‬# # #‫راج‬# # #‫سنة الخ‬# # #‫„ ال‬Steuerjahr“ steht. Der Sinn ist mehr oder weniger derselbe, insofern beide Interpretationen die ordnungsgemäße Verbuchung der eingenommenen Steuern bei der Steuerbehörde beinhalten. Zur Lesung des kursiv

96

Werner Diem

geschriebenen ‫ية‬# # # # # # ‫راج‬# # # # # # ‫ للخ‬sei noch angemerkt, daß sie durch die, was die Ligatur ‫ خراجـ‬angeht, entsprechende Schreibung von ‫ استحراج‬Z. 7 gestützt wird. 5 ‫ة‬# # #‫شمول‬# # #‫ م‬ist ein Zustandsakkusativ, der ‫فة‬# # #‫ري‬# # #‫يم الش‬# # #‫راس‬# # #‫ وردت امل‬Z. 3 aufnimmt. Die Form des initialen mīm ist typisch für den Duktus des Schreibens. šml I -hu bi- „umfangen mit; zuteil werden lassen“ wird in mamlūkischen Dokumenten speziell gebraucht, wenn eine Gnade zuteil wird, und zwar vor allem seitens des Herrschers, daneben auch anderer hochgestellter Personen, so hier in Z. 9. Siehe ABHeid II, Glossar s.r. šml. Hier ist der Umstand gemeint, daß den Erlassen die Signatur des Emirs zuteil geworden ist. 6 Der Ehrentitel ‫ة‬# #‫دول‬# #‫قعد ال‬# #‫„ م‬Sitz der Dynastie“ ist mir ansonsten nicht nachweisbar, aber ich sehe keine andere Lesung für den rasm ‫د‬#‫ع‬#‫ں‬# # # # # # ‫ م‬oder ‫د‬#‫ع‬#Ö# # # # # # ‫ م‬. Das ʿayn ist durch andere Schreibungen gesichert. – Die Lesung ‫ة‬#‫نوب‬#‫ ال‬ist gesichert, da ‫ة‬# # # # # # ‫اي‬#‫م‬#‫ع‬#‫س‬# # # # # # ‫ وت‬Z. 14 dieselbe Ligatur von wāw und initialem Häkchen in Form einer hohen Schleife aufweist. Das Wort erscheint in noch kursiverer Form in Z. 7 und 10. Daß ein mehrfach vorkommendes Wort zunächst einigermaßen deutlich und anschließend stark vereinfacht geschrieben wird, ist typisch für derartige Schriftstücke. Mit ‫ية‬# # # ‫لطان‬# # # ‫ة الس‬# # # ‫نوب‬# # # ‫ ال‬ist die Leibgarde des Herrschers gemeint. Während ihr Befehlshaber hier als ‫ية‬# ‫لطان‬# ‫ة الس‬# ‫نوب‬# ‫ير ال‬# ‫ ام‬erscheint, ist die übliche Bezeichnung ‫كبرى‬# # #‫ة ال‬# # #‫نوب‬# # #‫ راس ال‬oder ‫نوب‬# # #‫ة ال‬# # #‫وب‬# # #‫( راس ن‬siehe die Einleitung zum Stück). Eine Lesung ‫ راس‬statt ‫ر‬#‫ي‬# # # # # # ‫ ام‬ist aber nicht möglich, da der letzte Buchstabe ‫ ر‬und nicht ‫ س‬ist. Zur Identifizierung dieses Beamten siehe ebenfalls die Einleitung zum Stück. – ʾkd II ʿalā personae ist ein in Erlassen und verwandten amtlichen Schreiben üblicher Ausdruck, der allerdings mit fī rei konstruiert wird, während hier in der nächsten Zeile ‫ن‬# # ‫ وم‬. . . ‫ن‬# # ‫ م‬folgt. Nun könnte man sich zwar bei diesen beiden Schreibungen fragen, ob sie tatsächlich als ‫ن‬# # # # # # ‫ م‬und nicht vielleicht doch als ‫ فى‬zu lesen seien, aber das abschließende ‫ا‬#‫م‬# # # # # # ‫ وم‬zeigt, daß in der Tat ‫ن‬# # # # # # ‫ م‬gemeint ist. Die Lösung dieses Problems ist wie folgt. Die fī-Phrase von kd II ʿalā personae fī rei hat in Erlassen gewöhnlich die Form fī ḏālika (MSU S. XXXf.; AABW S.   18, Kommentar zu 2, 14f.). Es liegt deshalb die Annahme nahe, daß der Schreiber beim Zeilenwechsel ‫ك‬# ‫ فى ذل‬vergaß. Wenn man es dem üblichen Stil entsprechend ergänzt, dann können die min-Phrasen als Erläuterung des pauschalen ‫ فى ذلك‬aufgefaßt werden. Dementsprechend habe ich den Text ergänzt. 7 ‫فة‬# # ‫ري‬# # ‫ة الش‬# # ‫نوب‬# # ‫ور الى ال‬# # ‫ذك‬# # ‫ظ امل‬# # ‫مل على وش‬# # ‫ن الح‬# # ‫م‬: Zu ḥml I ʿalā „verfahren gemäß“ siehe AABW S. 215, Kommentar zu 44r, 8f. Der Kontext und die Konstruktion legen nahe, daß mit dem unpunktierten ‫ط‬# # # # # ‫ وس‬nicht ‫ط‬# # # # # ‫„ وس‬Mitte“, das hier überhaupt nicht paßt, sondern ‫ظ‬# # # ‫وش‬, Infinitiv von wšẓ I ʾilā „gehören zu“ gemeint ist.

Ein Begleitbrief von 904 H zu Erlassen aus dem mamlūkischen Ägypten

97

– ‫ه‬# # # # ‫حيطا ب‬# # # # ‫ م‬kann sich nicht auf die Erlasse beziehen, da diese einen Rückverweis als Fem. Sing. erfordern würden, sondern auf ‫فة‬# # # ‫ري‬# # # ‫ة الش‬# # # ‫نوب‬# # # ‫ور الى ال‬# # # ‫ذك‬# # # ‫ظ امل‬# # # ‫وش‬. Der Adressat soll jene Zugehörigkeit des Steuerpächters zur Kenntnis nehmen. Als jemand mit guten Beziehungen kann der Steuerpächter Vorzugsbehandlung erwarten. – Mit ‫ور‬#‫ذك‬#‫وان امل‬#‫دي‬#‫ ال‬wird auf die in Z. 4 erwähnte Steuerbehörde für Oberägypten verwiesen, in deren Zuständigkeit die Dörfer fielen. Zur Schreibung von ‫وان‬# # # # # # ‫دي‬# # # # # # ‫ ال‬sei noch angemerkt, daß der bestimmte Artikel in Form von zwei kleinen Häkchen geschrieben ist, wie es auch bei anderen Wörtern der Fall ist, und zwar vor allem dann, wenn der Beginn des Wortes hochgestellt ist. 8 Ergänzung von ‫ته املبلغ‬# # ‫ودت‬# # ‫ الى اخ‬nach Z. 5. – yuḥīṭu (oder tuḥīṭu) bihi lʿulūmu l-karīmatu: ähnlich wa-ṭuḥīti l-ʿulūmu š-šarīfatu ʾanna ABHeid I S. 93, Kommentar zu 16r, 7, ferner al-ʿulūmu l-karīmatu muḥītatun bi- APBW 47, 8f. – wa-lā ruḫṣata li-ʾaḥadin: ‫ ال‬ist etwas verkleckst, aber aus dem Kontext zu erschließen. Der Beginn von ‫د‬# # # # # # ‫ الح‬ist eigentümlich geschrieben, aber kaum anders zu lesen. Zu Ausdrücken mit ruḫṣah siehe AABW S. 321, Kommentar zu 74, 4. – muddah ist hier wohl als Zahlungsfrist zu verstehen. – ج‬# # ‫ امل‬ist aus dem Kontext ergänzt, der ein solches Wort verlangt. Es scheint, daß das Ende des Wortes ausgelassen, nicht verblaßt ist. Der Schriftzug ist wohl nachträglich hinzugefügt worden, da er rechts oberhalb von ‫ه‬# ‫دت‬# ‫ م‬steht, auf das er sinngemäß folgt. Es sind mir ansonsten nur Stellen mit al-marǧūwu mina llāhi bekannt; siehe ABW II S. 93 zu 14v, 2. In amtlichen Schreiben sind al-masʾūlu min und suʾāluhu min das Übliche; siehe zusammenfassend ABW II S. 114 zu 18, 11. 9 brz I wird typischerweise von Erlassen gebraucht; zu einigen Stellen siehe AABW, Glossar s.r. brz. – ‫ارات‬#‫ص‬# # # # # # ‫االق‬: ʾiqṣār steht hier statt des üblicheren taqṣīr. – naẓar ist hier und in Z. 10 „(wohlwollender) Blick“ im Sinne von Fürsorge. In Z. 10 ist naẓar mit ʿalā der begünstigten Person verbunden, wobei der Gedanke zugrunde liegt, daß der wohlwollende Blick „auf“ jemandem ruht. Eine andere Stelle dieser Art ist yakūn naẓaru l-maqarri l-ʾašrafi l-maḫdūmi ʿalayhi fa-l-mamlūku mamlūkuhu wa-taḥta naẓarihi l-karīmi APBW 47, 13f., wozu im Kommentar mehrere weitere Stellen angeführt sind. Das Wort ist in Z. 9 und an der ersten Stelle in Z. 10 korrekt mit ‫( ظ‬bzw. unpunktiertem ‫ )ط‬und anschließend an der zweiten Stelle in Z. 10 in vulgärer Schreibweise mit ‫( ضـ‬bzw. unpunktiertem ‫ )صـ‬geschrieben. – saʿīd ist im wesentlichen ein Epitheton ornans von Behörden, und es ist aus Sultansurkunden eine Behörde der Bezeichnung dīwān an-naẓar as-saʿīd bekannt (MSU S. 309 unten). Vielleicht war der Adressat der Angestellte

98

Werner Diem

eines solchen Ministeriums, so daß mit dem Epitheton eine Anspielung auf sein Amt erfolgte. 10 Das alif von ‫ة‬# # ‫ان‬# # ‫ االع‬ist nachträglich eingefügt worden. – In ‫نضر‬# # ‫ ب‬von ‫نضر‬# # ‫ب‬ ‫د‬#‫ي‬#‫ع‬#‫س‬# # # # # # ‫ ال‬ist das finale ‫ ه‬weggelassen, während es bei bei dem in derselben Zeile vorausgehenden ‫ نظره السعيد‬vorhanden ist. 11 ‫و‬# # # # # ‫ ه‬ist hier wie in Z. 7 stark vereinfacht geschrieben. – ‫ن‬# # # # # ‫عود م‬# # # # # ‫و امل‬# # # # # ‫ما ه‬# # # # # ‫ك‬ ‫ه‬# # # # #‫فضالت‬# # # # #‫ت‬: Statt ‫عود‬# # # # #‫امل‬, für das ich in solchem Kontext keine Stelle kenne, erwartet man das in gehobener Korrespondenz übliche stilistisch höherwertige ‫ود‬#‫ه‬#‫ع‬# # # # # # ‫امل‬. Siehe zu diesem Ausdruck ABHeid II 15, 6f. mit zusätzlichen Belegen im Kommentar. tafaḍḍulāt ist Plural des üblichen tafaḍḍul; dieser Plural ist mir ansonsten noch nicht begegnet. Dieselbe Form von ‫ مع‬findet sich in DF 21, 7 (S. 254), einem Erlaß von 851 H. – ‫ف‬# # # ‫ري‬# # # ‫مع التش‬: Ich nehme an, daß die unterbrochene lange Linie im zweiten Wort als eine einzige lange Linie gemeint ist. ‫ف‬#‫ري‬#‫ التش‬paßt; allerdings ist das fāʾ bei dieser Annahme sehr klein und wie ein Häkchen geschrieben. Der Text enthält keinen weiteren Fall von rechtsverbundenem finalem fāʾ, aber jedenfalls ist der Kopf von medialem fāʾ und qāf in diesem Dokument generell sehr klein geschrieben. Die Lesung ‫ف‬# # # ‫ري‬# # # ‫ التش‬liegt nahe, weil man hier einen Satz des Sinns erwartet, daß sich der Schreiber gegenüber dem Adressaten revanchieren werde. Die Präposition maʿa hat hier eine typisch epistolographische Funktion, bei der das in der maʿa-Phrase Ausgesagte in loser Form angeschlossen wird. Dieselbe Funktion hatte maʿa bereits in der viel früher üblichen Schlußformel maʿa l-kitābi bi-ḫabarika u.ä. (siehe hierzu ABHeid I S. 47, Kommentar zu 7, 6) gehabt. Daß man mit Briefen oder Anliegen des Adressaten „beehrt“ werden möchte, ist ein Topos des Briefes. Eine solche Stelle ist auch ABHeid II 47, 8f, und im Kommentar hierzu sind weitere Stellen dieser Art zitiert. Ergänzend seien noch zwei Stellen aus gehobener Stilistik angeführt: wamawlānā l-ʾamīru l-ǧalīlu ʿAḍudu d-Dawlati . . . walīyu mā yarāhu wa-yaʾmuru bihi . . . min tašrīfī bi-l-mukātabati wa-taṣrīfī fī ʿawāriḍi l-ḫidmatiʾin šāʾa llāhu „Unser Patron, der verehrte Befehlshaber ʿAḍud ad-Dawlah . . . hat die Macht, zu entscheiden und zu befehlen . . ., (nämlich) mich (weiterhin) mit Korrespondenz zu beehren und mit den dienstlich anfallenden Dingen zu betrauen, so Gott will“ Abū Isḥāq aṣ-Ṣābiʾ (st. 384 H): Muḫtār I S. 67, 2–4 / S. 99, 5–8 (Brief aṣ-Ṣābiʾs an den Būyiden ʿAḍud ad-Dawlah) und wa-li-l-ʾārāʾi l-mušriqati . . . fī qabūli hāḏihi l-ḫidmati wa-tašrīfi l-ʿabdi bi-l-ʾamṯilati l-maymūnati . . . li-yabḏula fī mtiṯālihā ġāyata ǧuhdih / wa-nihāyata wusʿih || mazīdu l-ʿulūwi wa-š-šarafi „Aber die strahlenden Meinungen . . . haben in Hinblick auf die Frage, ob dieser dienst-

Ein Begleitbrief von 904 H zu Erlassen aus dem mamlūkischen Ägypten

99

eifrige Brief akzeptiert und der Sklave mit glückbringenden Schreiben . . . beehrt wird, damit er sie mit aller Kraft befolge, welche ihm zur Verfügung steht, / indem er an die Grenze seiner Fähigkeit geht, || die größere Erhabenheit und Ehre“ al-Waṭwāṭ (st. 573 H): Rasāʾil I S. 19, 12–14 (Schreiben namens des Herrschers von Ḫuwārizm an den ʿAbbāsiden al-Muqtafī li-ʾamr Allāh). 11–12 wa-llāhu taʿālā yudīmu ʿalā l-ʿāliyati mawādda niʿamihi: Zu Eulogien des Typs wa-llāhu yudīmu ʿalā fulānin šayʾan siehe ABHeid I S. 307, Kommentar zu 69r, 17. Eine Eulogie mit mawādd ist lā ġayyara llāhu mā bihi min niʿamih / wa-lā qaṭaʿa minhu mawādda faḍlihi wa-karamih „Gott verändere nicht die Gnaden, die Er ihr gab, / und Er schneide sie nicht von den Elementen Seiner Gunst und Großmut ab“ Ibn Faḍl Allāh al-ʿUmarī: Masālik XII S. 136, 6f. / S. 179, 2f. (alQāḍī al-Fāḍil [st. 596 H] an al-ʿImād al-Iṣfahānī). – al-ʿāliyah steht hier elliptisch für al-ḥaḍrah al-ʿāliyah, was ein üblicher Ausdruck ist. 13 Das auf den Monatsnamen ‫دة‬#‫ع‬#‫ق‬# # # # # # ‫ذى ال‬, der kursiv ohne das finale ‫ ه‬geschrieben ist, folgende Graphem ist eine stark vereinfachte Schreibung des für diesen Monatsnamen üblichen Epitheton ornans ‫رام‬# # #‫الح‬. Stark, aber nicht ganz so extrem vereinfacht begegnet eine Schreibung dieses Epithetons zusammen mit dem Monatsnamen am Ende von DF 3 (S. 32), einem Erlaß von 790 H. 14 Das ʿayn von ‫ اربع‬ist in einem großen Bogen, dessen untere Hälfte durch eine Beschädigung verloren ist, mit dem ‫ و‬von ‫ة‬# # # #‫سعماي‬# # # #‫ وت‬verbunden. Da ‫ام‬# # # #‫ ع‬vorausgeht, wäre eigentlich die Femininform ‫ة‬#‫ع‬# # # # # # ‫ ارب‬zu erwarten, aber solche Fehler sind üblich. Das sīn von ‫ة‬# # # # # ‫سعماي‬# # # # # ‫ وت‬ist sehr kurz, was für Erlasse typisch ist, aber die Lesung ist gesichert, da das Wort von oben an das wāw angeschlossen ist, während initiales sīn im Duktus des Schreibens von unten angeschlossen ist. Die Lesung ‫ وسبعماية‬ist deshalb auszuschließen. 17 Zum Schlußzeichen siehe APBW S. 219, Kommentar zu 46, 23, ferner AABW und ABHeid II, jeweils Sachindex, s.v. „Schlußzeichen“.

100

Werner Diem

Literaturverzeichnis a) Mit Sigle zitiert AABW

Werner Diem: Arabische amtliche Briefe des 10. bis 16. Jahrhunderts aus der Österreichischen Nationalbibliothek in Wien. Wiesbaden 1996.

ABB

Werner Diem: Arabische Briefe des 7. bis 13. Jahrhunderts aus den Staatlichen Museen Berlin. Wiesbaden 1997.

ABH

Albert Dietrich: Arabische Briefe aus der Papyrussammlung der Hamburger Staats- und Universitäts-Bibliothek. Hamburg 1955.

ABHeid I

Werner Diem: Arabische Briefe auf Papyrus und Papier aus der Heidelberger Papyrus-Sammlung. Wiesbaden 1991.

ABHeid II Werner Diem: Arabische Briefe auf Papier aus der Heidelberger Papyrus-Sammlung. Heidelberg 2013. ABW II

Werner Diem: Arabische Briefe aus dem 10.–16. Jahrhundert. (Corpus Papyrorum Raineri 32.) Berlin 2011.

AGBW

Werner Diem: Arabische Geschäftsbriefe des 10. bis 14. Jahrhunderts aus der Österreichischen Nationalbibliothek in Wien. Wiesbaden 1995.

APBW

Werner Diem: Arabische Privatbriefe des 9. bis 15. Jahrhunderts aus der Österreichischen Nationalbibliothek in Wien. Wiesbaden 1996.

APEL

Adolf Grohmann: Arabic Papyri in the Egyptian Library. 1–6. Kairo 1934– 1962.

DAAF

Michele Amari: I diplomi arabi del R. Archivio fiorentino. Florenz 1863.

DF

Norberto Risciani: Documenti e firmani. Jerusalem 1931.

HJES III

Elihu Strauss-Ashtor: History of the Jews in Egypt and Syria under the Rule of the Mamluks. III. (Geniza Documents.) [Hebr.] Jerusalem 1970.

MSU

Hans Ernst: Die mamlukischen Sultansurkunden des Sinai-Klosters. Wiesbaden 1960.

RCEA XVII Répertoire chronologique d’épigraphie arabe. XVII. Ed. Ludvik Kalus. Kairo 1982.

b) Mit Verfasser und Kurztitel zitiert Abū Isḥāq aṣ-Ṣābiʾ: Muḫtār = al-Muḫtār min Rasāʾil ʾAbī ʾIsḥāq ʾIbrāhīm b. Hilāl b. Zahrūn aṣ-Ṣābiʾ. Ed. Šakīb Arslān. 1. Baʿabda 1898 / Beirut 2o. J. Bāšā: Alqāb = Ḥasan Bāšā: al-ʾAlqāb al-ʾislāmīyah fī t-tārīḫ wa-l-waṯāʾiq wa-l-ʾāṯār. Kairo 1957. Dozy: Dictionnaire = Reinhard Dozy: Supplément aux dictionnaires arabes. 1–2. Leiden 2. Aufl. 1881. Goitein: „Miktab“ = Shelomo Dov Goitein: „Miktab el ha-Rambam be-ʿinyene hahaqdašot wi-ydiʿot ḥadašot ʿal ṣeʾeṣaʾaw han-negidim.“ In: Tarbiz 34 (1965), S. 232–256.

Ein Begleitbrief von 904 H zu Erlassen aus dem mamlūkischen Ägypten

101

Ibn Faḍl Allāh al-ʿUmarī: Masālik XII = Ibn Faḍl Allāh al-ʿUmarī: Masālik al-ʾabṣār fī mamālik al-ʾamṣār. as-Sifr aṯ-ṯānī ʿašar. Frankfurt am Main 1988 / Ed. Ibrāhīm Ṣāliḥ. Abū Ẓaby 2002. Ibn Iyās: Badāʾiʿ III = Ibn Iyās: Badāʾiʿ az-zuhūr fī waqāʾiʿ ad-duhūr. III. Ed. Muḥammad Muṣṭafā. Kairo 32008. Ibn Taġrī Birdī: Manhal = Gaston Wiet: Les biographies du Manhal Safi. Kairo 1932. Halm: Lehensregister = Heinz Halm: Ägypten nach den mamlukischen Lehensregistern. 1–2. Wiesbaden 1979–1982. al-Qalqašandī: Ṣubḥ = al-Qalqašandī: Ṣubḥ al-ʾaʿšā fī ṣināʿat al-ʾinšā. 1–14. Beirut 1987. Ramzī: Qāmūs = Muḥammad Ramzī: al-Qāmūs al-ǧuġrāfī li-l-bilād al-miṣrīyah min ʿahd alqudamāʾ al-miṣrīyīn ʾilā sanat 1945. I. II 1–4. Indexband. Kairo 1953–1968. Sauvaget: „Noms“ = Jean Sauvaget: „Noms et surnoms de mamelouks.“ In: Journal asiatique 238 (1950), S. 31–58. al-Waṭwāṭ: Rasāʾil = Maǧmūʿat rasāʾil Rašīd ad-Dīn al-Waṭwāṭ. 1–2. Ed. Muḥammad Afandī Fahmī. Kairo 1315.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen. Versuch einer Typisierung1 Melanie Hanitsch, Universität Erlangen-Nürnberg

1. Einleitung In allen modernen arabischen Dialekten sind in mehr oder minder großer Zahl Verbalmodifikatoren anzutreffen, die vor die altererbte Perfekt- oder Imperfektform des Verbs treten und dessen temporale, aspektuelle und/oder modale Referenz eingrenzen oder spezifizieren. Ein Beispiel für die referenzielle Eingrenzung/Spezifizierung innerhalb des temporalen Bereichs ist die Partikel bā- (oder ba-, bi- b-), die sich in etlichen Dialekten von (im weitesten Sinne) nomadischem Typ findet. Für ihre Etymologie wurden verschiedene volitive Modalverben vorgeschlagen.   Sie tritt vor das morphologische Imperfekt (Präfixkonjugation), welches in den nomadischen Dialekten im Allgemeinen ein sehr weites Anwendungsspektrum als modusindifferentes Imperfektiv und „Non-Past“ – einschließlich futurischer Referenz – aufweist, und legt es in weitgehend eindeutiger Weise auf diese zuletzt genann1

In seiner Behandlung der modernen arabischen Dialekte nimmt Jan Retsö häufig eine weit gefasste vergleichend-typologische Perspektive ein (z.B. Retsö 1982–1983, 2004, 2006). Zudem hat er wichtige Arbeiten zur arabischen Verbalmorphologie verfasst und immer wieder auch die Diachronie der neuarabischen Tempus-, Aspekt- und Modalsysteme thematisiert (vgl. Retsö 1983, 1989, 2005, 2009, 2014). Der vorliegende Aufsatz möchte einen bescheidenen Beitrag zur weiteren Erforschung dieses faszinierenden Feldes beisteuern. Aufgrund des engen thematischen Bezuges und einiger inhaltlicher Überschneidungen, die zwischen diesem Aufsatz und meiner Dissertation bestehen (Verbalmodifikatoren in den arabischen Dialekten. Untersuchungen zur Evolution von Aspektsystemen), ist der Aufsatz als Teilpublikation/Vorabpublikation anzusehen. Für die Genehmigung zur Veröffentlichung danke ich dem Betreuer meiner Dissertation, Herrn Professor Werner Diem. Des Weiteren danke ich meinem Kollegen Salah A. Fakhry und seiner Schwester Israa Fakhri für die intensiven Diskussionen zum Bagdadischen, in denen die Idee zu diesem Beitrag wurzelt. Von Salah A. Fakhry stammen auch die mit BaM-Ṣ gekennzeichneten Beispiele im bagdadischen Dialekt.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

103

te Komponente, also die futurische Referenz fest,  2  z.B. iḫwānik iwaddūnik ‘your brothers will take you (there)’ (einfaches Imperfekt), gegenüber b-aštiġil ‘I shall work’3 (Imperfekt mit futurischem b-Präfix). Ein analoges Beispiel für den Bereich der aspektuellen Referenz bietet die Partikel ʾɛt- des Maltesischen, die sich sehr wahrscheinlich aus dem aktiven Partizip ʾiǝ:ʿɛd (< AA *qāʿid) des Verbs ‚sich setzen/sitzen, bleiben‘ herleiten lässt. Wird diese Partikel dem Imperfekt präfigiert, modifiziert sie dessen allgemein-imperfektivische Bedeutung (Ausdruck „allgemeingültiger Wahrheiten“, habitueller Sachverhalte und „aktuell vorliegender Sachverhalte“), z.B. (1) – (3), durch eine Festlegung auf den „aktuell vorliegenden Sachverhalt“. Die Partikel dient also der explizit-formalen Markierung einer Art von Progressiv, z.B. (4) – (5):4 (1) [Ma]  5 tāf li ʾattūs yislɔḥ fɛnɛk ‚Tu sais qu’un chat ça dépèce un lapin’ [Vanhove 1993: 74] (2) [Ma] sikwīt iʿīdu stɛyyɛr fūʾ-ʿa ‚Souvent on raconte des histoires à son sujet‘ [Vanhove 1993: 75] (3) [Ma] š inti taʿmɛl ɛmm fūʾ ‚Qu’est-ce que tu fais là-haut?’ [Vanhove 1993: 76]

2

Die Einschränkung „weitgehend eindeutig“ wird hier gemacht, weil Persson (2008: 35-37, 48-49) gezeigt hat, dass etwa das b-Imperfekt des Golfarabischen in bestimmten syntaktischen und textuellen Zusammenhängen auch andere Bedeutungen, vor allem die eines Irrealis, annehmen kann. Zur Etymologie sowie zur volitiv-futurischen (Grund-)Semantik des nomadischen bā-Imperfekts vgl. Kampffmeyer 1900: 53, Nöldeke 1904: 65-67, Cohen 1924: 64, Johnstone 1967: 143, 152, 163, 163, 169, Fischer – Jastrow 1980: 75, Rosenhouse 1984: 37, Persson 2008: 31 (unter Bezug auf Holes 2004: 247, Fn. 29 und Al-Tajir 1982: 80, 110) sowie Rosenhouse 2006: Bd. I, 266. Eine zusammenfassende Evaluation der Diskussion um den Ursprung verschiedener Ausprägungen des b(v)-Imperfekts, also auch einschließlich des b(i)-Imperfekts der sesshaften Dialekte, bietet Retsö 2014: 65-66.

3

Johnstone 1967: 152.

4

Vanhove bezeichnet die verschiedenen Bedeutungskomponenten der imperfektiven Domäne konkret als „[l]‘expression des vérités générales“, „[l]‘expression de faits habituels“ und „en situation de concomitance“ (Vanhove 1993: 73-75).

5

[Ma] = Maltesisch, [BaM] = Muslimisch-Bagdadisch, [Da] = Damaszenisch, [Dē] Dēri (Dēr iz-Zōr, Syrien), [MoC] = Christlich-Maṣlāwi (Mossul, Irak), [MoM] = Muslimisch-Maṣlāwi, [Š] = Šāwi-Dialekt von al-MoḤasan (Syrien), [Ṣa] = Ṣanʿānisch.

104

Melanie Hanitsch

(4) [Ma]

(5) [Ma]

yıǝ:n ʾɛt-iǧaʿlū-ni naʿmil-ʿa di-l-bičča šɔʿɔl ‚Moi, ils sont en train de me forcer de le faire, ce boulot‘ [Vanhove 1993: 115] il-baḥrīn ḥasbū-ḥ li ʾɛt-yɛʿrɛʾ ‚Les marins ont pensé qu’il était en train de se noyer’ [Vanhove 1993: 120].

Nur sehr wenigen neuarabischen Verbalmodifikatoren steht im Altarabischen ein klares formales und funktionales Pendant gegenüber, und noch seltener findet sich eine gesicherte etymologische Entsprechung. Da die Ausbildung der neuarabischen Verbalmodifikatoren zudem mit weitreichenden Umstukturierungen des arabischen TAM-Systems6 – in einigen Stadtdialekten sogar mit einer regelrechten Erneuerung desselben – verbunden ist,7 sind sie in Abgrenzung zum Altarabischen zu den wichtigsten morphosyntaktischen Charakteristika des Neuarabischen zu rechnen und verdienen entsprechende Aufmerksamkeit. In Übersichtsstudien zum Neuarabischen wird das Phänomen der Verbalmodifikation meist durch einzelsprachliche „Exemplare“ illustriert, die – wie etwa in den bereits angeführten Belegen – ihrer Form und Funktion nach als prototypische Vertreter ihrer Kategorie gelten können. Diese Prototypikalität residiert unausgesprochen in den drei folgenden Merkmalen: 1.) Der Modifikator ist partikelartig – sprich unflektiert und somit inkongruent – und oftmals von geringer phonologischer Substanz. 2.) Der Modifikator ist (nicht zuletzt für den Sprecher) etymologisch weitgehend intransparent. 3.) Der Modifikator weist eine relativ klar fassbare Beziehung zwischen Form und Bedeutung auf, die gut unter eine 1:1-Relation subsumiert werden kann. So kann z.B. das da-Imperfekt des Muslimisch-Bagdadischen in einer temporalen Perspektive als ein Präsens angesehen werden8 (1:1-Relation), bzw. in einer eher aspektuellen Perspektive als ein Progressiv9 (ebenfalls 1:1Relation); das bi-Imperfekt des Kairenischen ist in einer aspektuellen Per-

6

Das Akronym TAM steht im Folgenden für „Tempus, Modus/Modalität und Aspekt“.

7

Vgl. Czapkiewicz 1986: 43, Cohen 1984: 278-279, Simeone-Senelle 1985-1986: 58-59, Fischer 1982: 84, Jastrow 1982: 138, Eksell-Harning 1995: 66-67, 69.

8

Vgl. Jastrow 2007: Bd. II, 423.

9

Vgl. Erwin 1963: 139-140.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

105

spektive ein sehr allgemein gehaltenes (indikativisches) Imperfektiv (1:1Relation).10 In einzelsprachlichen Studien wird die in Punkt (3) angesprochene Beziehung zwischen der Form- und der Bedeutungsseite der Verbalmodifikation durchaus differenzierter betrachtet. So widmen sich einige Referenzgrammatiken sowie Schwerpunktstudien zu Tempus, Aspekt und/oder Modalität im Neuarabischen einer detaillierten Betrachtung der spezifischen sowie kontextabhängigen Lesarten der Verbalmodifikatoren. Sie befassen sich also gewissermaßen mit der Assymmetrie der Beziehung zwischen einer klar eingenzbaren Form (etwa bi- oder ʾɛt-) und ihren vielfältigen Bedeutungen.11 Weitaus seltener sind Untersuchungen anzutreffen, die sich mit Erscheinungen der Tempus- und Aspektmarkierung befassen, die hinsichtlich der Form-Bedeutungs-Relation in umgekehrter Weise asymmetrisch sind, so dass einzelsprachlich mindestens einer der folgenden Fälle zutrifft: a) Ein und dieselbe TAM-Bedeutung kann durch mehr als eine Konstruktion aus /VM + finitem Verb/ ausgedrückt werden; so z.B. im Golfarabischen, wo die Kategorie Futur sowohl durch das b-Imperfekt als auch durch das rāḥ-Imperfekt realisiert wird.12 b) Der Verbalmodifikator besteht aus zwei (oder mehr) Morphemen, die auch unabhängig voneinander als Verbalmodifikatoren fungieren, z.B. ḫal-dangūl ‚let’s say’ (Muslimisch-Bagdadisch), wobei sowohl ḫal(li) als auch da- in derselben optativen Bedeutung auch direkt und einzeln mit dem Imperfekt

10 Vgl. Cohen 1984: 286. 11 Cowell etwa unterscheidet beim b-Imperfekt des Damaszenischen vier Hauptfunktionen, „Future“, „Annunciatory“, „Generalizing“ und „Dispositional“ (Cowell 22005 (1964): 324-329. Eine Auswahl vergleichbar ausführlicher Kapitel in Referenzgrammatiken ist: Blau 1960: 84-101 (b-Imperfekt im Dialekt von Bīr Zēt), Reichmuth 1983: 286-299 (b-Imperfekt im Dialekt der Šukriyya, Sudan), Watson 1993: 7884 (bi-Imperfekt im Ṣanʿānischen), Woidich 2006: 278-280 (ḥa-Imperfekt im Kairenischen). Vergleichbare Abschnitte oder Kapitel in Detailstudien zum Aspekt finden sich in Eisele 1999: 90-95, 226-251 (bi-Imperfekt im Ägyptisch-Arabischen), Brustad 2000: 165-202 (dialektvergleichend; siehe insbesondere die Übersichtstafel S. 168; hier werden die Lesarten aufgeführt, die der „Indikativ des Imperfekts“ – darunter auch Formen wie das levantinische b-Imperfekt oder das marokkanische ka-Imperfekt – im Gebrauch mit unterschiedlichen semantischen Typen von Verben annehmen (Verben der Sinneswahrnehmung, Bewegungsverben, etc.), sowie Chaara 2003: 79-82 (ka-Imperfekt im Marokkanischen)). 12 Vgl. Persson 2008: 33-35.

106

Melanie Hanitsch

verbunden werden können.13 Als eine Steigerung dieses Falles kann (c) angesehen werden. c) Die syntagmatische Ausdehnung des verbmodifizierenden Komplexes (bzw. des TAM-Ausdrucks in seiner Gesamtheit) kann nicht ohne weiteres abgesteckt werden, z.B. yaʿnī ʾanā gad sirt atfarrajthā wa-ʿjabatnī gawī gawī / ʿinda-mā šahadt al-jamal ġarr jiss bī-dūr bī-dūr bī-dūr fī maḥallih / wa-yuxrujlak salīṭ ‚I mean, I have gone to watch it and I really enjoyed it when I saw the camel going round and round and round on the spot and you had oil coming out‘ (Ṣanʿānisch).14 Der vorliegende Beitrag ist vornehmlich mit dem Phänomen (b) befasst, das ich im Folgenden als „doppelte“ TAM-Markierung bezeichnen möchte. Die Phänomene (a) und (c) werden nur am Rande berührt, wenn sie einen direkten Beitrag zur Behandlung und zum Verständnis von (b) leisten. Darüber hinaus liegt der Schwerpunkt des Beitrags im Bereich der doppelten Aspektmarkierung, während auf die doppelte Tempus- und Modusmarkierung lediglich ein Ausblick geboten wird (s. Abschnitt 6). Im Bereich der doppelten Aspektmarkierung werden zwei bestimmte Formtypen einer schwerpunktmäßigen Betrachtung unterzogen: 1.) das /qāʿid Kv-Imperfekt/,15 oder auch „doppelte Sitzend-Imperfekt“, das im mesopotamischen Raum dokumentiert ist. Es setzt sich aus zwei Doubletten bzw. kognaten Elementen zusammen, nämlich aus dem jeweils lokalen Reflex des aktiven Partizips qāʿid ‚sitzend’ und einer formal reduzierten, partikelförmigen Ausprägung desselben, die in abstrakter Form als /Kv-/ dargestellt wird. Doppelte Aspektmarkierung aus Doubletten bzw. Kognaten bezeichne ich in Anlehnung an Georg Kampffmeyer (1900) auch als „pleonastisch“.16 13 Vgl. Blanc 1964: 116-117. 14 Watson 2000: 46/47; in diesem Beispiel wird die kontinuative bis progressivische Aspektnuance nicht durch das bi-Imperfekt allein ausgedrückt, sondern durch eine Reihe zusätzlicher Mittel gestützt/verstärkt: 1.) Durch die ikonische Wiederholung des finiten Verbs (vgl. Watson 2000: 46, Fn. 8); 2.) Durch den auf das Verb ‚sitzen‘ zurückgehenden Modifikator jiss (vgl. die volle Variante jilis) (vgl. Watson 1993: 158), sowie das Adverb ġarr (ġar ~ ġayr) ‚only, just‘ (vgl. Watson 1993: 334 und 1996: 302); zur aspektuellen bzw. „aspektverstärkenden“ Funktion von ġar(r) siehe auch das folgende Beispiel: ʾanā ġar bayn-aḥākīh w-hū gad fallatnī ‚I was just talking to him when he left me and went‘ (Watson 1993: 334). 15 Die Folge steht für /Konsonant + Vokal/. 16 Vgl. Kampffmeyer (1900: 57, Fn. 1) in Bezug auf eine „doppelte“ Futurmarkierung (s.u. Abschnitt 6.).

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

107

2.) das ʿam-b-Imperfekt, das sich, wenn man den gängigen Etymologien der beiden beteiligten Morpheme folgt, aus nicht-kognatem Material zusammensetzt.17 Diesen Typus der doppelten TAM-Markierung bezeichne ich daher in Abgrenzung zum zuvor genannten als „kumulativ“. Zwei Gründe sprechen für die schwerpunktmäßige Betrachtung ausgerechnet dieser beiden Formtypen: •

•

Die drei (etymologisch definierten) Modifikatoren (1) qāʿid/Kv-, (2) ʿam(māl) und (3) b- sind jeder für sich genommen – also zunächst ungeacht der Frage ihres Vorkommens im Rahmen einer doppelten Aspektmarkierung – in den modernen arabischen Dialekten sehr verbreit und weisen z. T. eine außerordentliche Formenvielfalt auf. Im Falle von ʿam(māl)- und b- weist auch die doppelte Aspektmarkierung selbst eine verhältnismäßig große geographische Verbreitung auf.

Diese beiden „quantitativen“ Gesichtspunkte, lassen eine typologische Betrachtungsweise überhaupt erst lohnend erscheinen. Die typologische Betrachtung wird im Wesentlichen geleitet von der Frage, wie sich die beiden Morpheme, aus denen sich in bestimmten Dialekten eine doppelte Markierung zusammensetzt, zueinander verhalten: Dieses „Verhalten zueinander“ wird sowohl innersprachlich als auch sprachvergleichend typisiert. Dabei lautet die innersprachliche Teilfrage: In welchem Ausmaß und unter welchen Bedingungen alterniert die einfache Aspektmarkierung (qāʿid oder Kv-, ʿam(māl) oder b-) mit der entsprechenden doppelten Aspektmarkierung (qāʿidKv-, ʿam(māl)-b-)? Die sprachvergleichende Teilfrage lautet: Wie sind Dialekte 17 Der Modifikator ʿam- und seine zahlreichen diatopischen Varianten gehen unzweifelhaft auf das auch heute vielerorts noch als Modifikator gebräuchliche Intensivnomen ʿammāl ‚tätig seiend‘ zurück (vgl. Lammens 1900: 559 [erwähnt in Stewart 1998: 104], Cohen 1924: 66-67, Fischer – Jastrow, 1980: 75, Agius – Harrak 1987: 168-172, Lentin 1994: 297-298, 300). Für den Modifikator b- sind die beiden dominanten Etymologievorschläge 1.) die konjunktional gebrauchte Präposition bi(vgl. u.a. Spitta 1880: 203, Cohen 1924: 64, Retsö 2014: 70), 2.) die im klassischen Arabisch belegte reduzierte Variante der Konjunktion baynā (< bayna-mā) (vgl. Behnstedt 1985: Bd. I, Karte 83 und zugehörige Anmerkung auf S. 22, 1987: 53, Fischer – Jastrow 1980: 75; kritische Diskussionen hierzu finden sich in Eksell 2005: 81 und Rubin 2005: 145-151). Weitere Etymologievorschläge leiten b- aus dem genannten ʿammāl her (vgl. Vollers 1887: 376, 393, erwähnt in Kampffmeyer 1900: 61), aus bada (AA badaʾa ‚anfangen‘) oder baqa (AA baqiya ‚bleiben‘) (vgl. Eksell 2005: 84-85) und schließlich auch, ebenso wie den „nomadischen b(ā)-Modifikator, aus diversen volitiven Modalverben, z.B. baġā (vgl. Eksell 1998: 70-71 und 2005: 82-83, letzteres unter Bezug auf einen Vortrag von Heikki Palva, Göteborg 1972).

108

Melanie Hanitsch

mit doppelter Aspektmarkierung in das Gesamt-Dialektgebiet eingebettet, in dem die beiden beteiligten Morpheme in irgendeiner Weise (also auch in Form von einfacher Aspektmarkierung) gebräuchlich sind? Dabei wird die typologische Aufstellung insbesondere auch auf ihre diachronischen Implikationen abgetastet: Es wird die Frage gestellt, welche Auskunft uns die einfache Aspektmarkierung mit ihrer Vielfalt an einzelsprachlichen Ausprägungen über die Entwicklung der doppelten Aspektmarkierung geben kann. 2. Formtyp doppeltes/pleonastisches Sitzend-Imperfekt Bevor wir uns der doppelten/pleonastischen Aspektmarkierung selbst zuwenden, soll ein kurzer dialektgeographischer Überblick über die Formenvielfalt der (auch und vor allem der „einfach“ gebrauchten) Sitzend-Modifikatoren gegeben werden. Die volle Form qāʿid (gāʿid, ǧāʿid, qēʿǝd, kāʿid) ist praktisch über das gesamte arabische Sprachgebiet verteilt. Besonders typisch ist sie hier für Dialekte nomadischen Typs oder mit nomadischem Adstrat. Darüber hinaus findet sie sich aber auch in Qǝltu-Dialekten des syromesopotamischen Raumes.  18 Die formal stark reduzierten, partikelförmigen Sitzend-Modifikatoren sind insgesamt vergleichsweise schwach verbreitet und weisen zwei regionale Schwerpunkte auf: 1.) Städte des mesopotamischen Raumes19 und hier insbesondere die alteingesessenen Communal dialects von Juden und Christen, welche dem Qǝltu-Typ angehören, 2.) in jüdischen Stadtdialekten des östlichen Maghrib sowie im Maltesischen, als dem insularen Ableger letzterer Dialektregion. Nachstehend wird der in der Literatur dokumentierte Bestand der reduzierten Formen im Detail aufgeführt:20 Irak: Das Jüdisch-Bagdadische hat den Modifikator qa-, mit dem volleren Allomorph qad- in der 1. Person Sg. In den Personen, deren Imperfekt-Personalpräfix auf t- oder n- anlautet, wird dieses bei Antritt des VM geminiert, z.B

18 Näheres zur geographischen Verteilung s.u. Abschnitt 2.2.1. 19 Vgl. Jastrow 2013: 5. 20 Formen, die vor allem durch meine eigenen Erhebungen dokumentiert sind, werden erst in Abschnitt 2.2.2.1. im Rahmen der typologischen Kontextualisierung vorgestellt.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

109

qattemši ‚she’s walking‘21 – ein Umstand, der natürlich auf eine Assimilation des auslautenden /d/ von qad- deutet und somit die Etymologie *qāʿid stützt. Das Christlich-Bagdadische hat im gesamten Paradigma einheitlich nur noch qa-;22 das Muslimisch-Bagdadische da-23 mit der nur noch sehr selten gebrauchten Variante ga-.24 In Mossul ist ein Modifikator belegt, der bis ins Detail dem des JüdischBagdadischen gleicht.25  Dieser ist jedoch möglicherweise einer lokalen, transkonfessionellen Koiné zuzuordnen bzw. sowohl unter Christen als auch unter Muslimen jeweils nur einer bestimmten Landsmannschaft/einem bestimmten

21 Die Gemination ist auf den Fall beschränkt, dass der VM in den Kontext {#_Kv} tritt, vgl. demnach qattemši ‚she’s walking‘ vs. qatġīd ‚she wants‘ (vgl. Blanc 1964: 115); die jüdisch-bagdadischen Texte von Mansour zeigen, dass die Gemination des Imperfektpräfixes (wohl aufgrund analogischen/paradigmatischen Ausgleichs) zudem auch auf die Formen mit dem Imperfektpräfix y- ausgeweitet wurde. Die Assimilation greift auch hier nur im o.g. Kontext, z.B. qayyǝtʿǝlmōn ‚[they] dy‘ (Mansour 1991: 68), aber qayʿǝllǝmǝm ‚er lehrt sie‘ (ebd. 1991: 68). 22 Vgl. Blanc 1964: 115, Abu-Haidar 1991: 88-89. 23 Zur formalen Ausprägung von da-: Stewart nimmt zur Erklärung des auslautenden /a/ eine Metathese an, die bei Übergangsformen wie qad- oder ʿad- (< qāʿid) ansetzt (vgl. Stewart 1998: 116). Mir erscheint es plausibler, der phonologisch schwachen 1. Person Sg. eine Schlüsselposition in der Herleitung dieser Form zuzumessen: *gāʿid_ámši trunkiert zu > d-ámši und reanalysiert als dá-mši, anschließend die Verallgemeinerung dieser Form durch paradigmatischen Ausgleich: da-timši, dayimši, etc. die schließlich auch die 1. Person Sg. selbst erfasst, wo die Variation damši ~ da-ʾamši vorliegt (Verbalformen angelehnt an Blanc 1964: 116). Rubin weist darauf hin, dass das Anhängen eines a-Vokals „curious“ sei, wofür er jedoch im ägyptisch-arabischen Futurpräfix ḥa- (< rāyiḥ) eine Parallele sieht (vgl. Rubin 2005: 137, Fn. 29). Es ist fraglich, inwiefern diese Parallelsetzung zulässig ist, denn Woidich führt ḥa- auf das Femininum der lexikalisch zugrundeliegenden Partizipialform rāyiḥ zurück, deren genusindifferenter Gebrauch in der Form raḥa- im Bʿēri (Oberägypten) belegt ist (vgl. Woidich 1995: 261). Für den genusindifferenten Gebrauch einer Femininform gāʿda liegt meines Wissens jedoch keine Evidenz vor. Eher scheint mir denkbar, auch die (scheinbare?) Genusverallgemeinerung von raḥa- zu einer Reanalyse der 1. Person Sg. in Verbindung zu setzen, in etwa rāyiḥ ʾamši > raḥ-amši reanalysiert als raḥa-mši und verallgemeinert zu raḥa-yimši. 24 Farida Abu-Haidar zufolge wurde die Form ga- in den neunziger Jahren nur noch von Personen im Alter von über sechzig Jahren gebraucht (bgl. Abu-Haidar 1994: 153). 25 Vgl. Blanc 1964: 116 und 196, Fn. 126, Jastrow 1979: 47.

110

Melanie Hanitsch

Subdialekt eigen.26 Informantenbefragungen und Textaufnahmen, die ich im Zusammenhang mit meinem Dissertationsprojekt mit drei christlichen und einem muslimischen Sprecher aus Mossul durchgeführt habe, bestätigen nämlich die v.a. in ältesten Quellen zu findenden,  27 aber auch durch Agius – Harrak (1987) bestärkten Hinweise28 auf zusätzliche, geringfügig anderslautende Modifikatoren in beiden Communal dialects: Das Christlich- Maṣlāwi hat ke- ~ kǝ- (seltener ka-) mit dem Allomorph ked(seltener kad-) vor der 1. Person. Dieses macht analog zum Formenbestand des Jüdisch-Bagdadischen und Muslimisch-Maṣlāwi (bzw. Koiné-Maṣlāwi) trotz des anlautenden Velars eine Herleitung aus *qāʿid wahrscheinlich. Zudem findet sich in Mossul ein muslimischer Subdialekt mit der Form ʿa(d)-, die als das Produkt einer Finaltrunkierung qāʿid > *ʿid- angesehen werden kann.29 Östlicher Maghrib: Partikelförmige und voll grammatikalisierte Modifikatoren aus *qāʿid finden sich in den Dialekten der Juden von Tunis (Tunesien) und Tripoli (Libyen), nämlich in beiden Fällen qa-,30 und – wie bereits eingangs erwähnt – auf Malta, nämlich ʾɛt-.31 Darüber hinaus weist in einigen Dialekten Tunesiens, die im Wesentlichen die volle Form qāʿǝd verwenden, dieses vor dem t- der 2. Person und der 3. Person f. Sg. eine reduzierte Variante qat- auf, z.B. winti ʾiš qattqūl? ‚Et toi qu’es-tu en train de dire?‘.32

26 Vgl. ähnliche Überlegungen in Jastrow 1979: 47, Fn. 33 sowie etwas allgemeiner auf die variationslinguistische Situation in Mossul bezogen Jastrow 2004: 139 (unter Bezug auf Blanc 1964: 10 und 184, Fn. 13). 27 Vgl. al-Čalabi 1935: 72 (Hinweis in Blanc 1964: 196, Fn. 126) und die Texte von Socin 1882 (vgl. Hinweis in Jastrow 1979: 47, Fn. 33). 28 Vgl. Agius – Harrak 1987: 165-166; Amir Harrak stammt selbst aus Mossul. 29 Zu diesem Formenbestand der beiden Communal (sub)dialects vgl. Agius – Harrak 1987: 165: „Within the city of Mawṣil and its surroundings, the Muslim and the Christian communities replace these three auxiliary variances [d.h. gāʿid, ǧāʿid und qāʿid] by ʿad-, qad- and ked- respectively”; zusätzliche innersprachliche wie sprachvergleichende Argumente für die Herleitung aller hier genannten Modifikatoren aus *qāʿid finden sich in Agius – Harrak (1987: 174-175, 177) sowie in meiner Dissertation. 30 Cohen 1975: 136-137 (erwähnt in Cohen 1984: 280), Yoda 2005: 193-194. 31 Vgl. Cohen 1984: 279, Vanhove 1993: 112. 32 Cohen 1984: 280.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

111

Nur in zweien der hier aufgeführten Dialekte ist eine doppelte Aspektmarkierung sicher dokumentiert, nämlich im Muslimisch-Bagdadischen33 und im Christlich-Maṣlāwi. 2.1 Innersprachliche Betrachtung der doppelten Aspektmarkierung Im Muslimisch-Bagdadischen und im Christlich-Maṣlāwi kann die einfache Markierung des Progressivs, also da- bzw. ke(d)-, z.B. (6) – (7), ohne Bedeutungsunterschied durch eine doppelte Progressivmarkierung ersetzt werden, z.B. (8) – (9):34 (6) [BaM] da-yuštuġuḷ bil-ḥadīqa ‚Er arbeitet gerade im Garten‘ [BaM-Ṣ] C (7) [Mo ] kedaštǝġǝl / kedasāwi kǝbbi kedasāwi dōlma ‚Ich bin am Arbeiten, ich bereite gerade Kubba und Dolma zu‘ [MoC-UA] M (8) [Ba ] gāʿid da-yuštuġuḷ bil-ḥadīqa ‚Er arbeitet gerade im Garten‘ [BaM-Ṣ] C (9) [Mo ] ʾaš kǝtǝʿmalīn… mǝn_aku ʿǝddki? – ʾǝtqūl: qaʿʿād kenǝštǝġǝ::l35 kenǝʿmal qahwi / ʿǝddna ḫǝṭṭār / ǧāna nēs ‚Was machst du gerade? Wer ist gerade bei dir? – Dann sagt sie: „Wir sind gerade beschäftigt, wir machen (gerade) Kaffee. Wir haben Besuch da… Es sind Leute (vorbei)gekommen“‘ [MoC-UA]. Für sich genommen gleichen sich die doppelte Aspektmarkierung des Christlich-Maṣlāwi und des Muslimisch-Bagdadischen. Der phonologisch vollere Modifikator qēʿǝd bzw. gāʿid, welcher im syntaktischen Verhältnis zum Nukleus der Verbalphrase als der „äußere“ Modifikator bezeichnet werden kann, ist semantisch ebenso verblichen, wie der „innere“ Modifikator, also die phonologisch reduzierte, präfigierte Doublette da- bzw. ke(d)-: Sie transportiert in dem betreffenden Zusammenhang keine konkrete Bedeutung des Sitzens mehr und ist dementsprechend, wie die Beispiele zeigen, auch auf Verben anwendbar, die stark agentivische, dynamische Handlungen bezeichnen. Diesen Grad der semantisch-funktionalen Grammatikalisierung bezeichne ich im Folgenden als Typ D-III, insofern als er die höchste/maximale Stufe (III) in der Ausbil33 Vgl. Fakhry 2015. 34 Bezüglich des Bagdadischen vgl. ebd. 2015. 35 Die Folge :: kennzeichnet eine prosodische Längung.

112

Melanie Hanitsch

dung/Grammatikalisierung einer doppelten Aspektmarkierung (D) repräsentiert.36 Auch hinsichtlich der Variation der doppelten Aspektmarkierung mit dem einfachen Kv-Imperfekt sind sich Muslimisch-Bagdadisch und ChristlichMaṣlāwi sehr ähnlich: Das progressivische /Kv-Imperfekt/ ist zwar bedeutungsgleich durch das /qāʿid Kv-Imperfekt/ ersetzbar. Doch ist „Progressiv“ nicht die einzige Funktion des einfachen Kv-Imperfekts. Vielmehr ist es in beiden Dialekten auch schon in unterschiedlichem Umfang in nicht-episodischer Lesart (z.B. Frequentativ, Habitualis) gebräuchlich,37 z.B. (10) – (11), obschon es in dieser Funktion bei weitem nicht obligatorisiert ist: (10) [BaM] da-yištaġlūn ib-jidd bil-mašrūʿ ij-jidīd ‚They’re working hard on the new project’ [Erwin 1963: 139]38 (11) [MoC] qǝttūlǝm: ǧīna hōni… baʿdēn nsāfǝr39 ʿal-Swēd… w-hōnīk40... bale ham wēḥǝd keyǝštǝġǝl… yaʿni bǝḏ-ḏahab bǝl-ǝSwēd / ham keyǝštǝġǝl / ʾē laqqālu ǧamāʿa qaʿʿad maʿāhǝm w-keyǝštǝġǝl ‚Ich sagte zu ihnen: Jetzt sind wir erst mal hierher gekommen (nach Syrien), als nächstes fahren wir dann nach Schweden. Dort habe ich auch einen [Sohn], der (dort) arbeitet… im Gold[schmiedehand36 Selbstverständlich ist das Konzept einer „maximalen“ Grammatikalisierung lediglich als Konstrukt anzusehen, das die typologische Einteilung erleichtern soll. Es liegt auf der Hand, dass Grammatikalisierung in dem Sinne „kein Ende hat“, insofern als auch einmal „voll“ ausgebildete grammatische Elemente (sozusagen Stufe III) einer weiter fortlaufenden Evolution unterliegen (vgl. Kurylowicz 1965: 52) und gegebenenfalls auch wieder vergehen (vgl. Lehmann 21995 (1982) 171-178). Jede Einteilung in Stufen, die sich an Konzeptionen von Minimum und Maximum orientieren, ist daher zwangsläufig arbiträr (vgl. die kritische Auseinandersetzung mit der „Unidirektionalitätshypothese“ und damit der Konzeption einer „steigerbaren“ Grammatikalisierung in Campbell 2001: 124-127), aber im vorliegenden Zusammenhang für eine erste Orientierung durchaus zweckmäßig. 37 Bezüglich des Muslimisch-Bagdadischen vgl. Erwin 1963: 139. 38 Erwins Transkription wurde unverändert übernommen, mit Ausnahme der Langvokale, die hier mit und wiedergegeben werden, statt durch und . 39 Vermutlich eine koineisierte Form, da in Mossul für Stamm III der Morphemtyp yfēʿǝl und zudem [ġ] statt [r] zu erwarten ist (vgl. Jastrow 1979: 38, 47). 40 Die Realisierung liegt zwischen hōnīk und hūnīk. Haim Blanc hat bei einem christlichen Sprecher aus Mossul hōnǝk, bei einem muslimischen Sprecher hingegen hnūka (vgl. Blanc 1964: 184, Fn. 13, zitiert in Jastrow 2004: 140).

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

113

werk]. Er hat auch Arbeit, ja, er hat dort ein paar Leute kennengelernt, mit denen er jetzt arbeitet (w. ließ sich nieder und arbeitet jetzt mit ihnen)‘ [MoC-UA]. In dieser letzteren Funktion kann das einfache Kv-Imperfekt nicht durch das qāʿid Kv-Imperfekt ersetzt werden. Es wird deutlich, dass der „äußere“ Modifikator gāʿid bzw. qēʿǝd das Kv-Imperfekt in derselben Weise modifiziert, also in seiner aspektuellen Referenz auf das Progressiv (bzw. temporal betrachtet auf das „aktuelle Präsens“) einschränkt, wie es in anderen Dialekten der einfache Modifikator (z.B. maltesisch ʾɛt-) mit dem einfachen Imperfekt tut (s.o.). Der Gebrauch des doppelten Aspektmarkierung scheint also ein klarer Fall von Verstärkung (reinforcement) zu sein, der im Grammatikalisierungsprozess dann einsetzt, wenn die Semantik des älteren Morphems in einem kritischen Maß verblichen und referenziell verallgemeinert ist.41 Der Unterschied zwischen der doppelten Aspektmarkierung der beiden Dialekte betrifft eine andere Dimension der „freien“ Variation: Im MuslimischBagdadischen ist auch die mit Hilfe der volleren Form gāʿid gebildete, einfache Aspektmarkierung uneingeschränkt gebräuchlich, z.B. (12). Auch sie entspricht in semantisch-funktionaler Hinsicht der Grammatikalisierungsstufe III, ist also Typ E-III (wobei E für „einfache Aspektmarkierung“ steht). Sie alterniert ohne Unterschied in der referenziellen Bedeutung mit den progressivischen Verbindungen des gāʿid da-Imperfekts und des da-Imperfekts: (12) [BaM] gāʿid yuštuġuḷ bil-ḥadīqa ‚Er arbeitet gerade im Garten‘ [BaM-Ṣ]. Im Christlich-Maṣlāwi hingegen ist dieser Formtyps der einfachen Aspekmarkierung nicht belegt. Das innersprachliche Nebeneinander der beiden Formtypen der einfachen Aspektmarkierung (gāʿid-Imperfekt und da-Imperfekt) im Muslimisch-Bagdadischen ist mit Blick auf den Grammatikalisierungsprozess ein klarer Fall von Schichtung (layering).42 Die in referenzieller Hinsicht bedeutungsgleiche Verwendung der doppelten Aspektmarkierung ist ein Fall von Verstärkung (rein41 Vgl. Lehmann 21995 (1982): 22-24. 42 Vgl. Hopper – Traugott 22003 (1993): 124-126, Bybee – Perkins – Pagliuca 1994: 2122.

114

Melanie Hanitsch

forcement) des da-Imperfekts, welches hierdurch explizit auf seine progressivische Kernbedeutung festgelegt wird; eine nicht-episodische Lesart hingegen wird ausgeschlossen. Im Christlich-Maṣlāwi sind die Verhältnisse dieselben, mit der Ausnahme, dass die ältere Schicht der beiden einfachen Aspektmarkierungen (qēʿid-Imperfekt) entweder schon aufgegeben wurde oder zumindest nicht mehr so produktiv ist, dass sie in spontansprachlichem Material auftauchte. Eine erneute Grammatikalisierung eines solchen „einfachen“ Sitzend-Imperfekts aus demselben Quellmaterial – die ja grundsätzlich denkbar ist – hat offenbar nicht eingesetzt. Neben der in semantischer Hinsicht voll grammatikalisierten Ausprägung von gāʿid da- (bzw. qēʿǝd ke(d)-) (also Stufe bzw. Typ D-III) liegen in beiden Dialekten jeweils auch formidentische Verbindungen vor, in denen das erste Element gāʿid bzw. qēʿid nicht oder kaum grammatikalisiert ist, z.B. (13)-(14). Es liegt also innersprachlich auch Schichtung (layering) zwischen der voll grammatikalisierten doppelten Aspektmarkierung und ihrer Quellkonstruktion vor. Die Unterscheidung der beiden Kontruktionstypen erfolgt rein auf der semantischen Ebene. Ein relativ einfach zu handhabendes erstes Kriterium zur Unterscheidung dieser beiden Typen ist, zu fragen, ob die betreffende Handlung im Sitzen erfolgt. Ist dies nicht der Fall, liegt sicher die voll grammatikalisierte Form vor (D-III). Ist die Antwort jedoch positiv, sprich, ist die Handlung eine solche, die üblicherweise im Sitzen erfolgt, genügt das noch nicht als Kriterium zum Ausschluss der „maximalen“ Grammatikalisiertheit: Das periphere Morphem gāʿid (bzw. qēʿid) kann durchaus trotzdem in seiner „voll“ grammatikalisierten Ausprägung vorliegen. Die Schwierigkeit der innersprachlichen Abgrenzung wird im Folgenden anhand des Chrislich-Maṣlāwi illustriert (14-16): (13) [BaM] šbīk intä gāʿid dätḥassib ‚Was hast du, dass du dasitzt und über finsteren Gedanken brütest‘ [Malaika 1963: 80] (14) [MoC] haḏōli [[qaʿʿād] keysōlfūn] ‚Diese sitzen da und unterhalten sich (wörtl. wobei sie sich unterhalten)‘ [MoC-UA] (15) [MoC] ʾana kǝntu kadaštǝġǝl b-bēt 43 / l-bēġḥa / [qēʿǝd [kadaktǝb]] hāyi lwaraqa44… mbōr ʾaġaššǝʿa l-Mīlāni 43 Die zu erwartende Form wäre bǝl-bēt ‚im Haus/zu Hause‘ mit bestimmten Artikel. Die Form b-bēt kann einer zufälligen Unterartikulation geschuldet sein. Vielleicht ist der Fall aber auch zu einer allgemeinen Tendenz der Qǝltu-Dialekte in Bezug zu

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

115

‚Ich war zu Hause „am Arbeiten“, gestern, ich war dieses Blatt „am schreiben“… um es Melanie zu zeigen‘ [MoC-A] (16) [MoC] [qaʿʿād ke-[nǝštǝġǝ::l]] ‚wir sind (gerade) am Arbeiten‘[MoC-A]. Üblicherweise wird angenommen, dass im Zuge der Grammatikalisierung eines Elements auf der syntaktischen Ebene eine Umkehrung der Dependenzverhältnisse stattfindet: das sogenannte rebracketing. 45 Während in (14) das vollwertige lexikalische Element qēʿǝd (Pl. qaʿʿād) ‚sitzend‘ durch einen asyndetischen Zustandssatz im ke-Imperfekt modifiziert wird (zentrale Klammer um qaʿʿād), steht im voll grammatikalisierten Fall (16) das einfache Imperfekt (imperfektiver Sachverhalt im weitesten Sinne) in der zentralen Klammer und wird durch den Komplex qaʿʿād ke- modifiziert. Dadurch wird der allgemein imperfektivische Sachverhalt als konkret zum Referenzeitpunkt vorliegend dargestellt. qaʿʿād ke- schränkt also die Referenz des Imperfekts explizit auf die progressivische Bedeutung ein. Beispiel (14) entspricht Stufe/Typ D-I (Quellkonstruktion der doppelten Aspektmarkierung), während Bsp. (16) der bereits weiter oben definierten Stufe/dem Typ D-III entspricht. Beispiel (15) schließlich ist als Repräsentant eines Übergangsstadiums anzusehen, insofern als hier die Dependenzverhältnisse ambig sind. Die plausibleste Art der Klammersetzung, die m.E. durch den weiteren (narrativen) Kontext nahegelegt wird, ist wie folgt: Die zentrale Klammer wird um das ka-Imperfekt als dem Nukleus des Prädikats angesetzt, während dieses wiederum durch qēʿid modifiziert wird. In (15) wird die von den Verhältnissen in (14) abweichende Klammersetzung – also der Shift des Nukleus vom Partizip zum finiten Verb – in erster Linie durch die Verschiebung des narrativen Fokus legitimiert bzw. suggeriert: In (14) wird vor allem eine Art Bild- oder Szenebeschreibung geboten. Es werden primär die zeitlich stabileren, konkret räumlichen Verhältnisse geschildert, setzen: Diese neigen dazu, den bestimmten Artikel in gewissen Kontexten zu elidieren, bzw. überhaupt den Status der Determination nicht konsequent zu kennzeichnen (vgl. Blanc 1964: 120, 125-128, Jastrow 1990: 71). 44 Zu gelegentlichem [r] statt [ġ] im Maṣlāwi vgl. Jastrow 1979: 38-39. 45 Vgl. Hopper – Traugott 22003 (1993): 51, 89; für den speziellen Fall der Reanalyse und des hierbei involvierten rebracketings der Verbindung /Partizip + Zustandssatz/ vgl. insbesondere Persson 2009: 265-269 in Bezug auf das Golfarabische; zahlreiche Beispiele für rebracketing liefert auch Woidich in seiner Übersichtsstudie zur Grammatikaliserung im Ägyptisch-Arabischen (vgl. Woidich 1995: 261-263).

116

Melanie Hanitsch

und diese sodann mit – gewissermaßen akzessorischen – Nebenaktivitäten (keysōlfūn) „gefüllt“. Anders in (15); hier etabliert schon die Form kadaštǝġǝl im vorangehenden Satz, die (im Übrigen sitzend stattfindende) Handlung als das eigentliche Thema. Auch der aspektuelle Charakter des Sachverhalts als Progressiv in der Vergangenheit (kǝntu kadaštǝġǝl) ist bereits durch grammatische Mittel explizit gemacht (das Auxiliar kān für die Vergangenheit, das ke- für den Progressiv). Das Partizip qēʿid tritt lediglich hinzu, um die nunmehr ihrerseits als akzessorisch gewerteten szenischen Rahmenbedingungen des Arbeitens zu spezifizieren. Dass der Fokus der Erzählung auf dem Arbeiten bzw. Schreiben liegt und auch verbleibt,46 zeigt v.a. auch die hier nicht gebotene Fortsetzung des Berichts, in der das Blatt erst verlorengeht und dann unter großem Aufwand wiedergefunden wird. Mit Blick auf die thematische Kontinuität und Relevanz von kadaštǝġǝl im Rahmen der geschilderten Episode stellt sich die Frage nach dem eigentlichen Zweck von qēʿid. Sicherlich besteht dieser nicht darin, dem Zuhörer unmißverständlich mitzuteilen, in welcher körperlichen Position das Schreiben stattgefunden hat (im Sitzen und nicht etwa im Gehen oder Stehen? Schließlich würde der Zuhörer kaum etwas anderes als die sitzende Position erwarten!). Vielmehr sind es lediglich die abstrakten Gegebenheiten, die im Umstand des Sitzens mitimpliziert sind, die für die Darbietung der Information von eigentlicher Bedeutung sind: die zeitliche Dauer, Konzentration, Mühe – und mit eben diesen Gegebenheiten klingen wesentliche semantische Elemente eines voll grammatikalisierten Progressivs bereits an, nämlich Durativität und ein vergleichsweise hohes Maß an Agentivität und Dynamik. Diesen Übergangstypus, der im Prozess der Ausbildung/Grammatikalisierung der doppelten Aspektmarkierung eine Zwischenstufe repräsentiert, kennzeichne ich im Folgenden als D-II. Was hinsichtlich der Klammersetzung die Diskrepanz zwischen Stufe/Typ D-II und D-III betrifft, sind folgende Faktoren von Bedeutung: Der „äußere“ Modifikator qēʿid ist semantisch so weit delexikalisiert und damit auch seiner konkreten ‚Sitzend‘-Bedeutung entledigt, dass er von ke- semantisch nicht mehr unterschieden werden kann, sondern zusammen mit diesem letzteren Element eine explizite, wenn auch redundante – oder eben „pleonastische“ – Markierung des Progressivs stellt.

46 Genauer gesagt auf dem Schreiben eines kleinen Glossars mit typisch maṣlawischen Vokabeln!

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

117

Nachdem für einen der beiden Dialekte, in denen eine voll grammatikalisierte „doppelte“ Aspektmarkierung nachgewiesen ist, auch ein Übergangstyp D-II (Bsp. 14) identifiziert wurde sowie die Kriterien explizit gemacht sind, die sich zu seiner Diagnose eignen, fällt auf, dass auch in Dialekten, in denen keine doppelte Aspektmarkierung im engeren Sinne belegt ist, neben den weitgehend ungrammatikalisierten (potentiellen) Quellkonstruktionen (Typ D-I) durchaus Konstruktionen anzutreffen sind, die als Übergangstypen im Sinne von D-II angesehen werden können, so etwa im Muslimisch-Maṣlāwi, z.B. (17) (= D-I) vs. (18) (= D-II): (17) [MoM] mā ʾana qēʿǝd ʿadassalla w-ʾatwannas ‚Ich sitze ja hier nicht nur herum und habe meinen Spaß und vergnüge mich‘ [MoM-AS] (18) [MoM] w-ʾana qēʿǝd ʿadaḥki ṣṣōt47 ʿāli… l-bēġḥa mǝn kān ǧāni talifōn mǝmBaġdād… mā ḥassētu ʿala nafsi ʿadaḥki b-ṣowt48 ʿāli! ‚Und ich habe die ganze Zeit ganz laut geredet… gestern, als ich den Anruf aus Bagdad bekommen habe. Ich habe überhaupt nicht gemerkt, dass ich [so] laut rede!‘ [MoM-AS].49 So ist also festzuhalten, dass sich im Muslimisch-Maṣlāwi die Ausbildung bzw. Grammatikalisierung einer doppelten Aspektmarkierung zumindest im Vollzug befindet. Das entscheidende Kriterium, das das Muslimisch-Maṣlāwi von der Ausbildung einer Stufe D-III trennt, ist der Umstand, dass trotz stark progressivischer Konnotation des qēʿid ʿa-Imperfekts, dieses offenbar semantischen Selektionsbeschränkungen unterliegt: Es ist nicht mit Verben belegt, die 47 Die zu erwartende Präposition b- ‚mit‘ fehlt. Stattdessen ist eine minimale Gemination des wortanlautenden /ṣ/ zu vernehmen. Dies lässt annehmen, dass die Präposition assimiliert wurde. Ähnliches liegt in den drei Communal dialects von Bagdad vor. Hier berichtet Haim Blanc von einer regelmäßigen Assimilation der Präposition b- an wortanlautendes /p/ und /f/, z.B. ppāra ‚worth one para‘ und ffels ‚worth one fels‘ (vgl. Blanc 1964: 121). 48 Die in Mossul zu erwartende Monophthongisierung (vgl. Jastrow 1979: 40) ist hier nur partiell vollzogen. 49 Eine Übersetzung, die sich stärker an der immer noch präsenten konkreten Bedeutung von qāʿid ‚sitzen‘ orientiert, wäre: ‚und ich habe die ganze Zeit dagesessen und ganz laut geredet‘, insofern als der Sprecher tatsächlich in einem Café saß, als er den Anruf erhielt. In jedem Falle bleibt jedoch die Andeutung einer kontinuativen bis progressiven Nuance bestehen, die in der Übersetzung durch ‚die ganze Zeit‘ wiedergegeben wird.

118

Melanie Hanitsch

Sachverhalte ausdrücken, die mit großer wahrscheinlichkeit nicht im Sitzen stattfinden, sondern im Stehen oder Gehen. So müssen wir in den Communal dialects der Städte des mesopotamischen Raumes mit Bezug auf die Ausbildung der doppelten Aspektmarkierung vorerst zwei Arten von Systemen unterscheiden: 1.) Systeme wie das Muslimisch-Bagdadische und das Christlich-Maṣlāwi, in denen die Grammatikalisierung der doppelten Markierung ihr vorläufiges Maximum erreicht hat und somit Schichtung (layering) zwischen den drei diachronischen Typen D I-II-III vorliegt – dabei ist das Christlich-Maṣlāwi als das progressivere dieser beiden Systeme anzusehen, da die doppelte Aspektmarkierung mit keiner „altertümlicheren“ Form der einfachen Aspektmarkierung (*qēʿid-Imperfekt) in Variation steht, sondern nur mit der formal evoluierteren (ke(d)-Imperfekt); 2.) Ein System, in dem bislang nur die Typen D I-II entwickelt sind. 2.2 Typologische Kontextualisierung und Diachronie Die Realisierung einer doppelten/pleonastischen Sitzend-Modifikation, wie wir sie in Abschnitt 1.1. gesehen haben, setzt natürlich voraus, dass der betreffende Dialekt überhaupt Doubletten eines Sitzend-Modifikators aufweist. Umgekehrt ist es jedoch keinesfalls so, dass alle Dialekte, die derartige Doubletten haben, zugleich auch Anzeichen für die Grammatikalisierung einer pleonastischen Aspektmarkierung aufweisen. Im Folgenden werden arabische Dialekte danach typisiert, in welchem semantisch-funktionalen Grammatikalisierungsgrad die Folge qāʿid + Imperfekt überhaupt vorkommt (Stufe/Typ E-I bis E-III), und ob das grammatikalisierte Sitzend-Imperfekt reduzierte Doubletten aufweist (qāʿid ~ Kv-). Wenn *qāʿid in Doubletten vorkommt, wird die Frage gestellt, unter welchen Bedingungen diese Doubletten in Juxtaposition stehen, sprich, wie weit gegebenenfalls die Folge qāʿid Kv-Imperfekt in dem betreffenden Dialekt grammatikalisiert ist (Stufe/Typ D-I bis D-III). 2.2.1

Nur phonologisch volle Form /qāʿid/ belegt

Dialekte, in denen die Verbindung qaʿid + einfaches Imperfekt ausschließlich Stufe E-I repräsentiert – also selbst ungrammatikalisiert ist und lediglich eine potentielle Quellkonstruktion für die Entwicklung eines grammatikalisierten Sitzend-Modifikators darstellt – ohne zugleich einen alternativen, aspektuellen Verbalmodifikator aufzuweisen, sind äußerst schwer zu finden. Sofern sich den Grammatiken überhaupt ausdrückliche Informationen zur „Nichtgrammatika-

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

119

lisiertheit“ des Sitzend-Partizips finden, hat der betreffende Dialekt in der Regel bereits einen anderen Verbalmodifikator, der (neben dem einfachen Imperfekt) zum Ausdruck des Progressivs/aktuellen Präsens verwendet werden kann. So stellt etwa Stefan Reichmuth für den Dialekt der Šukriyya im Sudan Folgendes fest: „Als Partizip in durativer Funktion, so wie es von Worsley (S. 37f.) und Trimingham (S. 71) mit Beispielen wie /gāʿid yaṭbuḫ/ „er kocht gerade“ für Omdurman beschrieben wird, ist gaʿad bei den Šukriyya jedoch nicht gebräuchlich.“50 Zum Ausdruck der „unmittelbaren, punktuellen Gegenwart“ wird stattdessen das einfache Imperfekt verwendet, z.B. hassaʿ / naktulhun / ʿēn / nōgid fī_srūj-hēlhun „‚Jetzt töten wir sie, (vor meinen) Auge(n) zünden wir die Sättel ihrer Pferde an!’ (Beschreibung einer Vision)“.51 Zum „zeitunabhängigen“ Ausdruck „andauernder […] Vorgänge“ schließlich wird das bi-Imperfekt verwendet, z.B. niḥna hāḏōl al-minanḏum maʿāk gabīlt-al-ḫalīfa lʿawaḍ ‚wir hier, die wir gerade mit dir sprechen, gehören zu al-Ḫalīfa l-ʿAwaḍ‘.52 Ähnliches trifft für die Beduinendialekte des Sinai zu. So erwähnt Rudolf de Jong kein in irgendeinem Umfang grammatikalisiertes Sitzend-Imperfekt. Dafür ist neben dem einfachen Imperfekt auch das aus den Sesshaftendialekten entlehnte b-Imperfekt gebräuchlich. Dieses dient durchaus auch zum Ausdruck eines „present continuous“,53 z.B. ya Ẓāhir, lēh ibtuẓrub… alʿasākir? ‚Oh Ẓāhir, why are you hitting… the soldiers?’ (ʿAgA)54 [ = Dialekt der ʿAgāylah].55 Für die Dialekte Algeriens findet sich der Hinweis auf ein (schwach) grammatikalisiertes Sitzend-Imperfekt ebensowenig. Dessen Stelle im TAM-System scheint in mehr oder minder starkem Umfang (s.u. Abschnitt 5.2.) von der präsentativen Partikel rā- + Personalsuffix‚ (< *sieh + mich/ihn…) ausgefüllt zu werden, deren eine Nebenfunktion es ist, „d’actualiser le procès“, sprich, den Sachverhalt zu einem Referenzzeitpunkt (z.B. Sprecherzeitpunkt) in Bezug zu setzen, was in funktionaler Hinsicht einem Progressiv oder Präsens gleichkommt.56

Überprüfen wir umgekehrt Dialekte, die ausdrücklich keinen anderen imperfektivisch-präsentischen Verbalmodifikator aufweisen (wie etwa die eher mit Sesshaften- und Stadtdialekten assoziierten Formen bi- oder ka-, und stattdessen das einfache Imperfekt als sehr allgemeines, indikativisches Imperfektiv 50 In grammatikalisierter Form ist nur das Perfekt von Sitzen belegt, insofern als gaʿad + Imperfekt „ingressive Funktion“ hat (vgl. Reichmuth 1983: 294). 51 Ebd. 1983 : 288. 52 Ebd. 1983: 287. 53 De Jong 2000: 318. 54 De Jong 2000: 319. 55 Vgl. De Jong 2000: 246. 56 Vgl. Cohen 1984: 281-282, Madouni 1999: 125, 132; z.B. huwwa rā-h yahdǝr mʿā-h u lbḥar bda ithǝwwǝl ‚Il était en train de discuter avec lui quand la mer commençait à s’agiter.’ (Sīdi Bel-ʿAbbès, Region Oran/Wahrān, Algerien), Madouni 1993: 125.

120

Melanie Hanitsch

verwenden, d.h. in ihrem TAM-System noch verhältnismäßig konservativ sind, zeigt sich im konkreten Einzelfall stets, dass die Verbindung ‚sitzend‘ + Imperfekt in zumindest leicht grammatikalisierter Form anzutreffen ist. Damit liegt in diesen Dialekten bereits mindestens Stufe E-II vor (s.u.). Dies bedeutet im Umkehrschluss, dass die in der Theorie als besonders altertümlich zu bewertenden Systeme, in denen zum Ausdruck des Progressivs (oder des aktuellen Präsens) keinerlei grammatikalisierte Modifikatoren vorliegen, sondern ausschließlich das einfache Imperfekt verwendet wird, heute kaum noch existieren. Vielmehr hat es den Anschein, als sei auch in den konservativsten Dialekten zumindest Schichtung (layering) zwischen Stufe E-I und E-II die Regel. Dies ist z.B. der Fall in den von Rosenhouse beschriebenen nordisraelischen Beduinendialekten: Hier ist das b-Imperfekt der unmittelbar benachbarten Sesshaftendialekte gänzlich ungebräuchlich. Statt dessen wird in nomadisch-konservativer Manier das einfache Imperfekt verwendet, z.B. winn išnū ʿyūnha tegādaḥ nār ‚how firey her eyes were! how her eyes were flashing like fire!‘57 Neben dem einfachen Imperfekt ist im Prinzip ein grammatikalisiertes gāʿid-Imperfekt vorhanden,58 allerdings deuten die Belege möglicherweise darauf, dass es vor allem auf Handlungen angewendet wird, die bevorzugt im Sitzen, oder zumindest in statischer bzw. Ruheposition ausgeführt werden, z.B. (huwwa) gāʿad yfukk arruzma ‚he is untying the package‘ (wahrscheinlich, aber nicht sicher, im Sitzen), mī gāʿada tunṭur han-nās ‚these people are waiting indeed‘,59 lummin wiṣal, mī ǧāʿda tunṭur ha-n-nās ‚when he arrived, there were these people sitting waiting’.60 Für die Entwicklungsstufe des TAM-Systems, in der die voll grammatikalisierte Form (Typ E-III) vorkommt, ohne dass die Typen E-I und E-II aufgegeben worden sind, lassen sich etliche Dialekte anführen.  61 Sie sind praktisch über das 57 Rosenhouse 1984: 119. 58 Vgl. Rosenhouse 1984: 37-38. 59 Ebd. 1984: 122. 60 Ebd. 1984: 119. 61 Es ist anzunehmen, dass die Asymmetrie in der Dokumentation der drei Stufen nicht nur auf die tatsächliche sprachliche Situation zurückzuführen ist, sondern auch einem Problem der Perspektive geschuldet ist. Selbstverständlich ziehen stark grammatikalisierte Elemente eher die Aufmerksamkeit des Linguisten auf sich und werden dementsprechend in Grammatiken eher erwähnt. Schwach grammatikalisierte Elemente oder bloße Quellkonstruktionen hingegen sind erst im Dialektvergleich erkennbar und werden entsprechend in einzelsprachlichen Untersuchungen

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

121

gesamte arabische Sprachgebiet verteilt und gehören entweder zum (im weitesten Sinne) nomadischen Typ, oder aber zu einer alten Schicht von Sesshaftendialekten, die sich durch die Bewahrung des aa. /q/ auszeichnen: ostarabische Dialekte (z.B. der Dialekt von Baḥrayn),62 Šāwi/ʿUrúbi-Dialekte des syrischmesopotamischen Raumes (z.B. der Dialekt von Kwayriš im Südirak),63 der Dialekt von Tunis (Muslime),64 Tschadarabisch,  65 Dialekte der Euphrat-Gruppe innerhalb des Qǝltu-Arabischen66 (z.B. das Ḫātūni, Syrien)67 und prähilalische Sesshaftendialekte Tunesiens 68 und Algeriens. 69 Beispiele für Stufe/Typ E-III sind: ams ana gāʿid wiyya dēlēn, gāʿid aḥišš ‚yesterday I was with them, threshing all day’ (Baḥrayn),70 huwwa w-martah ǧāʿdīn yišʿalūn bīh ‚er und seine Frau waren dabei, dort (Stauden) auszureißen‘ (Kwayriš, Irak),71 ʿwāsǝr lǝwwlād ma-qaʿda kān tǝddāwur ‚C’est congé; les enfants ne sont là qu’à tournailler‘72 (Takrouna, prähilalischer Stadtdialekt Tunesiens).73 Auch als Komplement zu Verben der sinnlichen Wahrnehmung – und damit zur Darstellung des Hintergrunds im Inzidenzschema – ist das voll grammatikalisierte gāʿid-Imperfekt anzutreffen, z.B. lumman ʾiǧa ʾabūha min ilḥaǧǧ ligāha ǧāʿida titwallad ‚als ihr Vater vom Pilgerzuge zurückkam, fand er sie genur selten erwähnt. Reichmuth (1983: 294) mit dem expliziten Anführen negativer Evidenz ist eine Ausnahme. 62 Vgl. Holes 2001: Bd. I, 432. 63 Vgl. Blanc 1964: 115-116, Denz 1971: 82, Fischer – Jastrow 1980: 75, Jastrow 1980: 155, Behnstedt 1997: Bd. I, Karte Nr. 161 („Verbmodifikator Präsens“) in Abgleich mit Karten Nr. 517-518 (Beduinendialekte). 64 Vgl. Singer 1984: 304. 65 Vgl. Abu-Absi 1995: 37. 66 Vgl. Talay 2011: 917; Behnstedts Dialektatlas nach zu schließen haben auch eine Reihe von Qǝltu-Dialekten, die dem anatolischen Zweig angehören, jedoch im äußersten nordöstlichen Zipfel Syriens gesprochen werden, den Modifikator qēʿid (vgl. Behnstedt 1997: Bd. I, Karte Nr. 161 („Verbmodifikator Präsens“) in Abgleich mit Karte Nr. 516 (Qǝltu-Dialekte). 67 Vgl. Talay 1999: 182. 68 Vgl. Cohen 1984: 279-280. 69 Vgl. Caubet 1996: 91 (unter Berufung auf eine unveröffentlichte Qualifikationsschrift von Jihane Madouni, 1991). 70 Holes 2001: Bd. I, 432. 71 Denz 1971: 82 72 Cohen 1984: 279. 73 Vgl. Singer 1980: 33.

122

Melanie Hanitsch

rade beim Gebären‘.74 Das einfache Imperfekt ist jedoch in diesem Kontext auch (nach wie vor) möglich, z.B. yā rēt ʿān ilʿurúbi tšūfni tiktil bīya ‚o wenn doch das Auge des Arabers sähe, wie du auf mich losschlägst‘.75 In einer Untergruppe von Dialekten weist das zum imperfektivischen Verbalmodifikator grammatikalisierte Sitzen-Verb eine besonders hohe syntagmatische Variabilität auf: Das Partizip + Imperfekt wird nur zum Ausdruck des Progressivs/Continuous (episodisches Imperfektiv) verwendet. Zum Ausdruck iterativer oder habitueller (nicht-episodischer) Imperfektivität, wird das finite Sitzen-Verb + Imperfekt verwendet. Dabei wird Habitualität in der Vergangenheit durch qaʿad + Imperfekt ausgedrückt, z.B. ana gaʿadt arūḥ is-safar awwal ‚I used to be constantly going on journeys abroad in the old days’.76 Nicht auf die Vergangenheit begrenzte Habitualität hingegen wird durch yiqʿad + Imperfekt ausgedrückt, z.B. tigʿad kill yōm triššha? ʿād mā tigʿad, ma yṣir ‚do you spray it (crop) every day without a break? If you don’t it won’t be any good‘.77 Diesen Systemtyp, in dem das Sitzen-Verb als imperfektives Auxiliar fungiert, das über seine eigene Flexion verschiedene Subtypen von Imperfektivität kennzeichnet, vertreten die ostarabischen Dialekte (s. Beispiele) und der Dialekt von Benghazi (Libyen).78

2.2.2 Doubletten belegt 2.2.2.1 Kontaktstellung der Doubletten ausgeschlossen In diese Gruppe gehören der Šāwi-Dialekt von al-MoḤasan (Syrien), einer kleinen Ortschaft am Mittleren Euphrat, ca. 12 km südöstlich von Dēr iz-Zōr, sowie der Qǝltu-Dialekt von Dēr iz-Zōr selbst. Dem Belegmaterial in einschlägigen Grammatiken nach zu urteilen, gehören wohl auch das Maltesische mit den Formen ʾiǝ:ʿɛd und ʾɛt-,79 sowie der jüdische Dialekt von Tunis mit qāʿǝd ~ qa-80 hierher. Um dies jedoch mit Sicherheit sagen zu können, wäre die Berücksichtigung größerer Textcorpora und die Erhebung negativer Evidenz bezüglich der Kontaktstellung vonnöten, wie dies im Falle des Šāwi-Dialekts und des Dēri im Rahmen meines Dissertationsprojektes erfolgte. Daher möchte ich mich im vorliegenden Abschnitt auf diese beiden letztgenannten Dialekte konzentrieren. Hinsichtlich des Ausprägungsgrades der Sitzend-Modifikation ist der ŠāwiDialekt von al-MoḤasan geringfügig konservativer als das Dēri, was zunächst 74 Denz 1971: 82. 75 Ebd. 1971: 76. 76 Holes 2001: Bd. I, 432. 77 Ebd. 2001: Bd. I, 432. 78 Vgl. Owens 1984: 155-156. 79 Vgl. Vanhove 1993: 112. 80 Vgl. Cohen 1975: Bd. II, 136-138.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

123

nicht überrascht, insofern als man von Nomadendialekten a priori mehr Konservatismus erwartet als von Sesshaftendialekten, zu denen das Dēri als QǝltuDialekt ja zu rechnen ist. In der Gesamtschau frappiert allerdings vor allem die große Ähnlichkeit, welche zwischen den Aspektsystemen dieser beiden – vor allem genetisch, aber auch „grobtypologisch“ – so verschiedenen Dialekte besteht. Diese Ähnlichkeit ist sicherlich auf die große geografische Nähe der beiden Ortschaften sowie auf die enge soziolinguistische Verbindung zwischen den beiden Sprechergemeinschaften zurückzuführen, die ja auch für den ausgeprägten Mischcharakter des Dēri als Sesshaftendialekt mit nomadischem Adstrat verantwortlich zu machen ist und damit letztlich für die „feintypologische“ Nähe des Dēri zu den umgebenden Nomadendialekten.81 Diese Ähnlichkeit in Form und Systematik gilt nicht zuletzt auch für den Sitzend-Modifikator. Im Šāwi-Dialekt hat der Sitzend-Modifikator die Varianten bzw. etymologischen Doubletten gāʿid (ǧāʿid) ~ gāʿd und gaʿd-. Im Dēri lauten sie qēʿid ~ qaʿd-. In beiden Dialekten ist das Sitzend-Imperfekt überwiegend als Typ E I-II anzutreffen. Typ E-III – sprich die Anwendung des Modifikators auf Handlungen, die eine konkret sitzende Position ausschließen oder zumindest sehr unwahrscheinlich erscheinen lassen – ist nur sehr marginal belegt, z.B. (24) [Š] und (28) – (29) [Dē] und wird von den Sprechern z. T. auch als stilistisch grenzwertig beurteilt.82 Im Šāwi ist die volle Form gāʿid als Typ E-I bis III belegt (s. Bsp. (19) – (24)). Ab Stufe E-II bleibt sie in einem Teil des Paradigmas obligatorisch unflektiert, z.B. (20).83 81 Vgl. Jastrow 1980: 202, Behnstedt – Woidich 2005: 105. 82 So z.B. empfanden die Gewährspersonen beider Dialekte die Anwendung des Sitzend-Imperfekts auf die Translokationsverben ‚hinaufgehen’ und ‚hinuntergehen’ als grammatisch grenzwertig. Speziell im Dēri wurde sein Gebrauch ausdrücklich nur in Kombiantion mit Modaladverbien als grammatisch bewertet, z.B. qaʿd-itḥaddir šwayy šwayy (aber: ?qaʿd-itḥaddir#). Verben, die eine konkret sitzende Position ausschließen oder unwahrscheinlich machen und dennoch mit dem Sitzen-Imperfekt (Typ III) spontan vorkamen, waren in beiden Dialekten von nichtagentivischem Charakter: sterben, ertrinken, zittern, sieden, überkochen. Bei anderen sowohl dynamischen als auch agentivischen Verben wie ‚schlagen’ oder ‚schlachten’ wurde der Gebrauch des Sitzend-Imperfekts durch den Šāwi-Sprecher klar abgelehnt. 83 Gänzlich ausgeschlossen ist, der direkten Elizitierung nach zu urteilen, die Bildung des f. sg., sowie die Anwendung des pluralischen Partizips auf die 1. Person Plural. In der 3. Person pl. m. liegt allem Anschein nach Variation vor.

124

Melanie Hanitsch

Die beiden reduzierten Varianten gāʿd- und gaʿd- sind nur in Stufe E-II belegt, also in dem Funktionsbereich, in dem das grammatikalisierte SitzendImperfekt vergleichsweise uneingeschränktest verwendet wird (Bsp. (21) – (23)). Dabei ist die am stärksten reduzierte Variante, nämlich gaʿd-, auf den Gebrauch im Hauptsatz beschränkt und pragmatisch stark markiert. Ihr Gebrauch weist eine gewisse funktionale Überschneidung mit der in Nordafrika verbreiteten deiktisch-aspektuellen Partikel rā- + Personalsuffix auf. So kennzeichnet sie einen indirekten Sprechakt – genaugenommen einen indirekt ausgedrückten Direktiv (explizit: Du VERB-st ja! – implizit: lass das!), z.B. (22), oder stellt die implizite Antwort auf einen expliziten Direktiv dar (explizite Antwort: Ich VERB-e doch gerade/schon! – indirekte/implizite Antwort: ‚deswegen kann/ brauche ich das, wozu du mich aufforderst, nicht (mehr) tun!‘), z.B. (23):  84

84 Fadoua Chaara zufolge belegt die Partikel rā das Prädikat des Satzes mit dem Satzfokus, und stellt die Satzaussage insgesamt als besonders „wahrhaftig“ und „informativ“ dar. Bei der Übersetzung ins Deutsche wird rā gerne durch die Fokus-/Gradpartikeln ‚ja‘, ‚bestimmt‘, ‚doch‘, ‚schon‘ wiedergegeben (vgl. Chaara 2003: 102-106). Die Satzaussage wird also zu einer im Kontext präsenten Alternativmenge in Kontrast oder Bezug gesetzt (vgl. Lemma „Gradpartikel“ in Bussmann 32002: 258, sowie Madouni (1993) in Bezug auf rā- im Dialekt von Sīdi Bel-Abbès (Region Oran/Wahrān, Algerien): „La particule rā permet […] à l’énonciateur un recentrage sur une valeur qu’il a choisie et qui existait antérieurement“ (Madouni 1993: 133), z.B. rā-k tšūf, hæ hiyya ṭōmāṭés ‚Tu vois bien, non, les voilà les tomates!’ (sur un ton quelque peu exaspéré […]) Madouni 1993: 130. Gerne bestehen die „Alternativmengen“, zu denen das mit rā fokussierte Prädikat in Kontrast gesetzt wird, in einem noch nicht realisierten Sachverhalt (auf den durch einen impliziten oder expliziten Direktiv Bezug genommen wird) oder in einem schon realisierten Sachverhalt (Bezugnahme durch eine Absage an einen impliziten oder expliziten Direktiv), z.B. rah l-makla sxuna ġer sir ġref u-kul ‚das Essen ist (doch) warm; geh hol dir […] etwas für dich und iss (Chaara 2003: 103)‘, oder u-zaʿma ra-k ntina ʿarfa šinni kayn bla ma n-ʿawed-lek ‚ich glaube du weißt doch [schon] alles, ich brauche es dir nicht zu erzählen (ebd. 2003: 104)‘. Die maghrebinische Partikel rā hat je nach Dialekt mehr oder minder stark ausgeprägte Sekundärfunktionen als Verbalmodifikator des Imperfekts, genauer gesagt, als Marker des progressiven Aspekts (z.B. in Algerien, vgl. Boucherit 1987: 12-18). In den maghrebinischen Dialekten mit der Partikel rā und im Šāwi-Dialekt mit seiner verbalmodifizierenden Partikel gaʿdliegt also eine jeweils spiegelbildlich gelagerte „Durchlässigkeit“ zwischen der Domäne des Imperfektivs/Progressivs (Aspekt) einerseits und dem Bereich der Fokus-/Gradmarkierung (Informationsstruktur) andererseits vor.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

125

(19) [Š] Typ I: [[gāʿdīn] nitfarraǧ ʿat-talfizyōn] ‚Wir sitzen hier und sehen Fern (w. wobei wir fernsehen)’[Š-M] (20) [Š] Typ II (lang): ʿiddna ǧabas / [ǧāʿid [nākul ǧabas]] ‚Wir hatten Wassermelone. Wir waren gerade dabei Wassermelone zu essen’ [Š-AA] (21) [Š] Typ II (mittel): š-gām itsawwi? – ʾagullu „gāʿd_atfarraǧ ʿat-talfizyōn“ ‚Was machst du da? – [da] sage ich zu ihm: ich schaue gerade fern’ [Š-M] (22) [Š] Typ II (kurz) : yaʿni šābb w-gāʿid ʿal-bāb / yidaḥḥig ʿar-rāyiḥ w-iǧ-ǧāy […] yaʿ# tuʿtubur ʿēb / ʾiǧīt ʿalē / š-gaʿd issawwi ʿal-bāb? š-gaʿd issawwi? ‚Da ist ein Junge, der sitzt vor der Haustür und schaut den Leuten nach, wie sie kommen und gehen […] Sowas tut man nicht ! Ich gehe also zu ihm hin [und sage]: ‚He, was machst du denn da, was machst du denn da?! (d.h. das tut man nicht!)’ [Š-M] (23) [Š] Typ II (kurz): gūm sawwīlna kāsit šāy – ʾagullu : y-aḫi gaʿd-anawwim Barhūm » ‚Komm, mach uns mal ein Glas Tee ! – Da sage ich zu ihm : Du siehst doch, dass ich gerade den kleinen Ibrāhīm in den Schlaf wiege (d.h. ich kann jetzt nicht !)’[Š-M] (24) [Š] Typ III : šūfu l-wlēd gāʿid yiġrag ‚Seht mal, der Junge ist am Ertrinken!‘ [Š-M]. Die formale Ambiguität (Typ E-I bis E-III) der Verbindung gāʿid + Imperfekt, deren erstes Glied im einen Extremfall als Vollverb, im anderen Extremfall als „maximal“ grammatikalisierter Verbalmodifikator interpretiert werden kann, wird gelegentlich aufgelöst, indem zwischen Partizip und Imperfektform ein anderer Verbalmodifikator, nämlich gām,85 inseriert wird, z.B. (25).86 Dadurch wird die in dieser Verbindung verblichene lexikalische Bedeutung des Partizips gestärkt. Es liegt also ein etwas spezieller Fall der Verstärkung (reinforcement) vor, der nicht auf den Erhalt der Bedeutung eines grammatikalisierten Elements, sondern eines ungrammatikalisierten/rein lexikalischen Elements abzielt und

85 Vgl. die in Behnstedts Sprachatlas von Syrien dokumentierte parallele Form qām- ~ qam- (aus AA *qāma ‚aufstehen‘) im Dēri (vgl. Behnstedt 1997: Bd. I, Karte Nr. 161 [„Verbmodifikator Präsens“]). 86 Damit liegt hier der Form nach eine potentielle Quellkonstruktion für die Grammatikalisierung einer nicht etwa „pleonastischen“, sondern vielmehr „kumulativen“ (weil auf der Juxtaposition von nicht-kognatem Material basierenden) Ausprägung der doppelten Aspektmarkierung vor. Die analoge Verbindung /gāʿid gaʿd-Imperfekt/ ist im Šāwi hingegen ungrammatisch.

126

Melanie Hanitsch

genau damit aber eine neue Konstruktion schafft, die sich einer erneuten Grammatikalisierung anbietet. (25) [Š] Typ D-I: ʾāni [[gāʿid] gām-ašrab DēqahwaDē] ‚Ich sitze da, und trinke Kaffee’87 [Š-M]. Der wesentliche Unterschied zwischen dem Šāwi und dem Dēri besteht darin, dass die Kurzform im Dēri eine geringfügig breitere Verteilung und eine erkennbar weiter gefasste Bedeutung hat. Zum einen ist die Kurzform auch in Typ E-III anzutreffen, z.B. (29), zum anderen unterliegt sie nicht den für die Variante gaʿd- des Šāwi aufgezeigten pragmatischen Restriktionen. Außerdem kann es ebenso wie das parallel gebräuchliche Stehen-Imperfekt (qām-Imperfekt) auch in nicht-episodischer Lesart, nämlich als Frequentativ, verwendet werden. Im Šāwi ist diese letztere Funktion noch gänzlich dem Stehen-Imperfekt (gāmImperfekt) vorbehalten. Ein eher konservativer Zug des Sitzend-Imperfekts des Dēri, in Abgrenzung zum Šāwi, besteht darin, dass die volle Form des Modifikators in allen drei Konstruktionstypen (E-I bis E-III) volle Genuskongruenz aufweisen kann, z.B. (28): (26) [Dē] (27) [Dē] (28) [Dē] (29) [Dē]

Typ II (lang): walla qāʿid yidrus ‚Ja, er ist gerade am Lernen’ [Dē-H] Typ II (kurz): ʾana qaʿd-aḥči maʿāči_ assaʿ ‚Ich rede gerade mit dir (wobei wir hier sitzen)’ [Dē-M] Typ III (lang): qāʿde tmūt w-trīd itšūf ʾibinha ‚Sie liegt im Sterben und möchte ihren Sohn sehen’ [Dē-H] Typ III (kurz): qaʿd-itfūr il-qidriyye / qūmi ʿalēha ‚Der Topf kocht gleich über (w. ist gerade dabei überzukochen)! Geh (schnell) hin!’ [Dē-H].

Eine Besonderheit beider Dialekte, Šāwi und Dēri, besteht darin, dass sich der intermediäre Grammatikalisierungsgrad von Sitzend + Imperfekt, also Stufe E-II, in der Elizitierung durch objektivere Kriterien diagnostizieren lässt als etwa die analoge Stufe D-II im Muslimisch-Maṣlāwi. Während im Maṣlāwi nur 1.) die Frage, ob das Subjekt konkret sitzt, sowie 2.) die (kon-)textuelle Ein87 Zur Disambiguierung gegenüber ʾāni gāʿid_ašrab, das, auch als ‚ich trinke gerade’ verstanden werden kann.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

127

bettung einen Rückschluss auf das Vorliegen einer (Übergangs-)Stufe II zulassen, kann diese im Dēri und Šāwi dadurch eindeutig identifiziert werden, dass sie jederzeit durch das fast bedeutungsgleiche, aber etwas weitläufiger gebrauchte Stehen-Imperfekt paraphrasiert werden kann, z.B. ʾiḥna qām-niḥči ‚wir reden gerade‘ [spontane Reformulierung mit Wechsel der Person:] ʾana qaʿdaḥči ‚ich rede gerade‘ [Dē-H]. 2.2.2.2 Kontaktstellung der Doubletten möglich Eine Kontaktstellung zweier Varianten bzw. etymologischer Doubletten von qāʿid ‚sitzend‘ ist, wie bereits in Abschnitt 1.1. ausgeführt, im MuslimischBagdadischen (gāʿid da-) und in mindestens zwei Communal dialecs von Mossul möglich, nämlich im Muslimisch-Maṣlāwi (qēʿǝd ʿa-) und Christlich-Maṣlāwi (qēʿǝd ke-). Dem Formenbestand der involvierten Dialekte nach zu urteilen besteht eine zentrale Voraussetzung für diese dritte Stufe in der Entwicklung der SitzendModifikation darin, dass die phonologische Erosion der partikelförmigen Modifikatoren nunmehr so weit vorangeschritten ist, dass im Bewusstsein der Sprecher keinerlei Gefühl für den Bezug zum Vollverb ‚sitzen‘ mehr vorhanden ist: Muslimisch-Bagdadisch da-, Christlich-Maṣlāwi ke(d)- und subdialektales Muslimisch-Maṣlāwi ʿa(d)-. Im Rahmen der fomalen Reduktion ist es vermutlich der Verlust des wortanlautenden Konsonanten (##da-, ##ʿa(d)-) bzw. seine phonologische Verschiebung qa(d)- > ka(d)-, der das entscheidende Moment für den Verlust der etymologischen Transparenz darstellt, insofern als generell der Wortanlaut für die Worterkennung von ungleich höherer Bedeutung sein dürfte als andere Bestandteile. Die Kappung des erkennbaren etymologischen Bezuges zwischen den vollen und den reduzierten Formen schafft offenbar die Voraussetzung dafür, dass eine Kontaktstellung zwischen ihnen beiden nicht mehr als redundant empfunden wird. Dies eröffnet der reduzierten Form zwei verschiedene Wege der Ausbreitung in einen neuen Kontext, nämlich in den Zustandssatz nach dem aktiven Partizip ‚sitzend’ (Vollverb). Der partikelförmige VM kann also eindringen in Konstruktionen von Typ E-I /qāʿid-Imperfekt/ ‚er sitzt, wobei er VERB-t‘, so dass wir eine zunächst synonyme Konstruktion des Typ D-I erhalten, also /qāʿid Kv-Imperfekt/ ‚er sitzt, wobei er VERB-t’. Der erste Ausbreitungsweg beruht auf der bereits im Šāwi in marginaler Weise dokumentierten Strategie, die Konstruktion /qāʿid-Imperfekt/, die im Zuge der Grammatikalisierung von qāʿid ambig geworden ist (E-I, II, oder III)

128

Melanie Hanitsch

durch eine explizite morphosyntaktische Markierung zu disambiguieren: Das Šāwi hat, wie bereits gezeigt, die Möglichkeit, den Status von gāʿid als Vollverb ‚sitzen’ zu stärken, indem es dem Imperfekt des Zustandssatzes eine eigenständige Progressivmarkierung gām präfigiert. Ähnliches finden wir auch im Bagdadischen. Jedoch kann die Verstärkung oder Renovation hier – vermutlich dank der etymologischen Intransparenz von da- – mit Hilfe des kognaten Aspektmarkers erfolgen, z.B. (30): (30) [BaM] šbīk intä gāʿid dätḥassib ‚Was hast du, dass du dasitzt und über finsteren Gedanken brütest‘ [Malaika 1963: 80]. Der zweite anzunehmende Ausbreitungsweg, der da- in diesen Kontext einführt, beruht auf einem Analogieschluss. Wir haben gesehen, dass in etlichen nomadischen Dialekten, die zwar keine doppelte Aspektmarkierung, aber ein voll grammatikalisiertes Sitzend-Imperfekt (E-Typ III) aufweisen, dieses auch in Komplementsätzen nach Verben der sinnlichen Wahrnehmung gebräuchlich ist, wobei es hier vermutlich in situ grammatikalisiert ist88 (s.o. Abschnitt 1.2.1.). Im Bagdadischen finden wir vergleichbare Konstruktionen, z.B. dawwarit ʿala ʾabūy / ligēta gāʿid yuštuġul bil-ḥadīqa ‚ich suchte nach meinem Vater (und) fand ihn im Garten am Arbeiten‘ [BaM-Ṣ] Dabei kann anstelle von gāʿid, dem formalen Evolutionsgrad dieses Modifikators entsprechend und ohne Unterschied in der referenziellen Bedeutung, auch da- stehen, also ligēta da-yuštuġul bil-ḥadīqa [BaM-Ṣ]. Vermutlich ist es nun wiederum die etymologische Intransparenz von da-, die es diesem Element ermöglicht, per Analogie auch im Zustandssatz nach Partizipien von Positionsverben zu stehen – und damit auch in Kontaktstellung zum ungrammatikalisiertem gāʿid ‚sitzend’, vgl. ligā da-yuštuġul ‚er fand ihn am arbeiten‘ und gāʿid da-tḥassib (s.o.). Ist mit Typ D-I ‚er sitzt, wobei er VERB-t‘ die Kontaktstellung zwischen gāʿid und da- erst einmal grundsätzlich zulässig, ergibt sich für die Konstruktion gāʿid-Imperfekt in kontrastiver Abgrenzung ein allmählicher Rückgang ihrer Interpretierbarkeit als synonymes E-I ‚er sitzt, wobei er VERB-t‘: Die Verbindung qāʿid + Imperfekt wird zunehmend als mindestens E-Typ II ‚er ist am VERB-en (wobei er sitzt)’ oder sogar E-Typ III ‚er ist am VERB-en’ interpretiert. Im zeitge-

88 Also ‚er fand ihn (da)sitzend, und am VERB-en‘ reanalysiert als ‚er fand ihn am VERB-en‘.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

129

nössischen Bagdadischen sind diese beiden letzteren Interpetationen die einzig zulässigen, gāʿid yuẓrub_aḫmās ib-asdās kann also nur interpetiert werden als ‚er ist ins Grübeln vertieft/am Grübeln’ und nicht als ‚er sitzt da und grübelt‘ [BaMṢ]. Im Muslimisch-Maṣlāwi ist der Prozess der funktionalen Ablösung und damit letztlich die Verdrängung von Typ E-II (*qēʿid-Imperfekt) durch Typ D-II (qēʿid ʿa-Imperfekt) möglicherweise abgeschlossen. Zumindest findet sich keine positive Evidenz für ihren Erhalt. Dasselbe gilt für das Christlich-Maṣlāwi, wo Typ D-II (qēʿid ka-Imperfekt ‚er ist am VERB-en [im Sitzen]‘ zudem zu Typ D-III weiter grammatikalisiert ist (qēʿid ka-Imperfekt ‚er ist am VERB-en‘). Positive Evidenz für einen parallelen Erhalt von formal nicht reduzierten aber voll grammatikalisiertem E-Typ III (*qēʿid-Imperfekt ‚er ist am VERB-en‘) wie er im Muslimisch-Bagdadischen noch weithin gebräuchlich ist, liegt auch hier nicht vor. Ein letzter Faktor, der sich vermutlich ab einem gewissen Punkt der Entwicklung auf die Verdrängung des E-Typs aus den Stufen I-III besonders förderlich auswirkt, hängt mit der zweiten Komponente des E-Typs, nämlich dem „einfachen“ oder „nackten“ Imperfekt und seiner Stellung im TAM-System insgesamt zusammen. Schon Haim Blanc (1964) bemerkte, dass der Gebrauch des Bagdadischen da-Imperfekts zur Kennzeichnung von „present time and noncontingency“ weniger stark verallgemeinert zu sein scheint, als der Gebrauch von kognatem qa- und qa(d)-Imperfek im Jüdisch- und Christlich-Bagdadischen.89 Hinsichtlich des Grades der semantisch-funktionalen Verallgemeinerung des Verbalmodifikator-Imperfekts stehen das Muslimisch-Maṣlāwi und das Christlich-Maṣlāwi eher den beiden letzteren Varietäten nahe, so dass in dieser Frage innerhalb der Tigris-Gruppe des Qǝltu-Arabischen eine recht große Einheitlichkeit zu bestehen scheint. David Cohen (1984) klassifiziert die drei Communal dialects von Bagdad als repräsentativ für jene neuarabischen Dialekte, in denen innerhalb des Indikativs des Imperfekts ein „dédoublement aspectuel“ vom Typ „concomitant“ (Verbalmodifikator-Imperfekt) vs. „non-concomitant“ (einfaches Imperfekt) vollzogen ist.90 Mit Blick auf den naturgemäß vereinfachenden Charakter einer jeden Typologie ist allerdings festzuhalten, dass sich die hier besprochenen Qǝltu-Dialekte der Tigris-Gruppe bereits deutlich im Übergang zu jenem anderen Typus von TAM-System befinden, welcher in 89 Blanc 1964: 116. 90 Vgl. Cohen 1984: 287-288.

130

Melanie Hanitsch

Cohens Typologie durch das Kairenische mit seinem bi-Imperfekt und das urbane Marokkanische mit seinem ka-Imperfekt repräsentiert wird, und wo das Verbalmodifikator-Imperfekt weitgehend zu einem indikativischen Imperfektiv verallgemeinert ist, wodurch das einfache Imperfekt wiederum weitgehend auf den nicht-indikativischen Gebrauch eingeschränkt ist.91 In einem System mit einem schon weithin als Indikativ gebräuchlichen Verbalmodifikator-Imperfekt wird nun ein asyndetisch angeschlossenes einfaches Imperfekt im Kontrast hierzu von den Sprechern leicht als modal, möglicherweise auch final reanalysiert werden. Sobald also ein Sachverhalt ausgedrückt werden soll, der faktisch und zeitgleich mit dem durch das vorangehende Verb ausgedrückten Sachverhalt vorliegt, müsste das einfache Imperfekt vermieden werden: *qēʿid-Imperfekt ‚?er sitzt, um zu VERB-en? > qēʿid ka-Imperfekt ‚er sitzt, wobei er VERB-t‘. 2.2.2.3 Kontaktstellung der Doubletten obligatorisch? Dialekte, in denen gāʿid-Imperfekt und Kv-Imperfekt nicht mehr unabhängig voneinander als Progressiv (oder in anderen imperfektivischen Funktionen) vorkommen, sondern nur noch als Komponenten einer „doppelten“ Aspektmarkierung auftreten oder gar zu Portmanteau-Morphemen verschmelzen, sind mir nicht bekannt. In dieser Leerstelle im System liegt ein augenfälliger Unterschied zu jenem anderen Formtyp der doppelten Aspektmarkierung vor, nämlich zur „kumulativen“ Aspektmarkierung durch das ʿam-b-Imperfekt und seinen vielfältigen formalen Varianten, die nun in Abschnitt 2 betrachtet werden sollen. 3. Formtyp doppeltes/kumulatives ʿam-b-Imperfekt Anders als im Falle der doppelten/pleonastischen Aspektmarkierung durch qāʿid Kv-, die nach meinem Kenntnisstand bisher nicht eingehend und um ihrer selbst Willen untersucht wurde, ist die grundsätzliche Möglichkeit der Verbindung von ʿam- und b- zu einer doppelten Markierung /ʿam-b-/ in der Literatur im Prinzip bekannt.92 Dieser Umstand, zugleich aber auch die weite Verbreitung und vergleichsweise hohe Diversität in der einzelsprachlichen Ausgestaltung dieses „kumulativen“ Formtyps, lassen es – anders als im Falle von /qāʿid

91 Vgl. Cohen 1984: 284, 286. 92 Vgl. Lentin 1994, sowie sämtliche in Abschnitt 3.1.2.2. verwendete Literatur.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

131

Kv-/ – angebracht erscheinen, die innersprachliche Betrachtung93 der doppelten Aspektmarkierung direkt in die „typologische Kontextualisierung“ zu integrieren. Die Frage nach der diachronischen Interpretierbarkeit der typologischen Daten – in diesem Falle ein ungleich schwereres Unterfangen, das nach weiteren, vertiefenden Untersuchungen verlangt – wird hingegen in einen gesonderten, zusammenfassenden Abschnitt ausgegliedert (s. Abschnitt 3.2.). 3.1 Typologische Kontextualisierung 3.1.1 Dialekte, die nur eines der beiden Morpheme aufweisen Dialekte, in denen bi- vorliegt,94 aber kein ʿam(māl), sind vor allem in der Peripherie des bi-Verbreitungsgebietes zu finden, so etwa im bereits erwähnten Dia93 Für /qāʿid Kv-/ erfolgte sie in Abschnitt 2.1. 94 Im vorliegenden Abschnitt sind diejenigen Dialekte berücksichtigt, deren b(v)Imperfekt der folgenden Mischung aus formalen und semantischen Kriterien genügt: 1.) Der Modifikator hat die form b(i)- oder zumindest ein so lautendes Allomorph und ist außerhalb der 1. Person Sg. nicht in der Folge [baK] anzutreffen; 2.) Sein semantisch-funktionaler Schwerpunkt liegt im Bereich des „aktuellen Präsens“ bzw. „Progressivs“ oder „imparfait concomitant“. Dieses an Kampffmeyer angelehnte Kriterium (vgl. Kampffmeyer 1900: 54), ermöglicht eine Abgrenzung gegenüber dem vor allem in Dialekten nomadischen Typs (oder mit nomadischem Adstrat) verbreiteten b(a)-Modifikator. Dessen Kernfunktionen sind in weitgehender Harmonie mit seiner als wahrscheinlich anzusehenden Entwicklung aus einem Verb des Wollens in erster Linie modal (insbesondere volitiv) bis futurisch (s.o. Fn. 2, sowie Kampffmeyer 1900: 53-59 und Retsö 2014: 65-67). Der im Kern volitiv-futurische Modifikor b(a)- ist über den Gebrauch in Konditionalsatzgefügen zuweilen zum Irrealis weitergrammatikalisiert und reicht sogar bis in die „Peripherie“ der Imperfektivität hinein. Das kann im Einzelfall durchaus zu Abgrenzungsschwierigkeiten zwischen dem hier behandelten “präsentisch-imperfektivischen” b(i)-Imperfekt und dem eher im arabischen Südosten gebräuchlichen „modal-futurischen“ b(a)-Imperfekt führen. Man beachte hierzu die folgenden Beispiele aus dem Golfarabischen: al-walad iḏa b-izzawwaǧ maṯalan zoǧteh tgūl… ‚if a boy is getting married, his (espoused) wife will say…’ (Persson 2008: 36), an-nās byikūnu ʿāyšīn fi ḫiām ‚(If I had lived 20 years ago) people would be living in tents‘ (ebd. 2008: 36), ʿad[d]ad sukkān ʿUmān kull yōm b-izīd ‚The number of inhabitants in Oman will increase/increases every day’ (ebd. 2008: 42; insbesondere zur Frage der Habitualität siehe Perssons Interpretation in ebd. 2008: 44-48). Auf die Frage, inwieweit einzelne der im vorliegenden Abschnitt berücksichtigten b(i)-Imperfekta (insbesondere aus der geographischen Peripherie des b(i)-Verbreitungsgebietes) möglicherweise bei einer detaillierten Analyse ihrer Funktion und

132

Melanie Hanitsch

lekt der Šukriyya im Sudan (s.o. Abschnitt 2.2.1.), in Nigeria95 oder im jemenitischen Hochland,96 z.B. keyf bi-tsabbir? ‚wie machst du das?‘ (Saʿdah).97 Doch auch im syropalästinischen Raum sind derartige Systeme in nicht-städtischen Sesshaftendialekten („Bauerndialekten“) anzutreffen, so etwa in den Dialekten der Dörfer um Ramallah (Palästina/Westjordanland)98 einschließlich Bīr Zēt,99 z.B. šu biddak. kāl: badawwir ʿassarband wilward ‚was willst du? – Ich suche nach S. und W.‘ (Bīr Zēt);100 vermutlich im Dorfdialekt von Abu Shusha (ehemals bei Distribution am Ende doch dem b(a)-Imperfekt zugeschlagen werden müssen, sei hier nur am Rande verwiesen. Ebenso bleibt die sehr komplexe Fragestellung, ob in einzelnen Teilregionen des (mutmaßlichen) b(i)-Verbreitungsgebietes in Wirklichkeit Kontamination und Zusammenfall von b(i)- und b(a)-Paradigmen vorliegen könnte, einer gesonderten Untersuchung vorbehalten. 95 Vgl. Owens 1993: 105-110. 96 Das Verbreitungsgebiet umfasst das nördliche Hochland mit der Stadt Saʿdah (vgl. Diem 1973: 29-33, Behnstedt 1985: Bd. I, Karte Nr. 83 und 169, Retsö 2014: 64), das zentrale Hochland einschließlich der Hauptstadt Ṣanʿāʾ (vgl. Behnstedt 1985: Bd. I, Karte Nr. 83 und 169, Retsö 2014: 64), sowie die Region al-Maḥābšeh (vgl. Diem 1973: 73), als dem (sprachlichen) Übergangsgebiet zwischen der nördlichen Hochebene und der Küstenebene (vgl. ebd. 1973: 71). Vereinzelt ist der Modifikator biaber auch im südlichen Hochland anzutreffen, für das ansonsten der Modifikator ḏī- oder ḏā- typisch ist. So hat etwa Ḏamār, ebenso wie Ṣanʿāʾ, die Form bi- mit dem Allomorph bayn- in der 1. Person Pl. Dies unterstreicht die auch ansonsten deutliche sprachliche Mittelposition dieser Ortschaft zwischen den ländlichen Dialekten der südlichen Hochebene und dem Ṣanʿānischen (vgl. Diem 1973: 45-61, insbesondere 50, 53). 97 Behnstedt 1987: 53; die übrigen Dialekte der Region um Saʿdah haben überwiegend die volle Form, bayn-, bēn, bīn-, oder bin-. Zwei Orte der Region, nämlich im-Ṭalḥ und Sāgēn stellen hinsichtlich des Grades der formalen Reduktion des Modifikators, ähnlich wie das Ṣanʿānische, einen Übergangstypus dar: Die 1. Person weist noch die volle Form bayn- bzw. bēn- auf, während der Rest des Paradigmas bi- hat (vgl. Behnstedt 1987: 53, auch erwähnt in Eksell-Harning 2005: 81 und Rubin 2005: 146; Piamenta 1991: Bd. I, 18, 40, 47-48, erwähnt in Stewart 1998: 120). Die Frage, inwiefern die Etymologie baynā ~ bayna-mā, die sich somit für das jemenitische bi-Präfix aufdrängt, für das gesamte Verbreitungsgebiet des imperfektivisch-präsentischen b-Präfixes gelten kann, oder ob für andere Teilgebiete andere Etymologien angenommen werden müssen (vgl. die anderen Etymologien in Fn. 17), soll an dieser Stelle nicht vertieft werden. 98 Vgl. Seeger 2013: Bd. III, 181. 99 Vgl. Blau 1960: 78, 143; den offenbar vereinzelt belegten Modifikator ʿammāl bewertet Blau als Entlehnung aus dem Ägyptischen. 100 Blau 1960: 78.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

133

Ramle),101 im Ḥōrān (Syrien/Jordanien)102, höchstwahrscheinlich in der Oase Soukhne103 und vermutlich in einer älteren Stufe des Dialekts von Tadmur/ Palmyra.104 Exponenten des b(i)-Modifikators, die nicht nur peripher sind, sondern vom Hauptverbreitungsgebiet in der einen oder anderen Weise isoliert stehen, sind das p(i)-Präfix im arabischen Dialekt von Kormatiki (Zypern)105 und das m(i)-Präfix im Dialekt von Bukhara (Usbekistan).106

101 Vgl. Kimary 2000: 31-33. 102 Vgl. Cantineau 1946: 221-222. 103 Behnstedt erwähnt in seiner Grammatik des Dialekts von Soukhne zum Ausdruck der „aktuellen Handlung“ sowie der „habituelle[n] oder sich wiederholende[n] Handlung“ nur das b-Imperfekt (vgl. Behnstedt 1994: Bd. II, 184-185). Der Dialektatlas von Syrien (vgl. Behnstedt 1997: Bd. I, Karte 160) erweitert dieses Bild um weitere „präsentische“ Modifikatoren (siehe weiter unten in diesem Abschnitt), jedoch nicht um ʿam(māl). 104 Cantineau (1934: 132-135) erwähnt nur das b-Imperfekt, wobei nicht ganz klar wird, inwiefern es zum Ausdruck eines „aktuellen Präsens/Progressivs/imparfait concomitant“ gebraucht werden kann: „L’infectum à b-préfixé sert fréquemment à exprimer, soit un fait d’ordre général, vrai en tout temps, soit une action que l’on fait tous les jours habituellement […] Mais pour exprimer le présent proprement dit: l’action ou l‘état au moment present, on a tendance à employer (au moins pour certains verbes), non pas l’imperfectum à b-préfixé, mais les participes […].” (Cantineau 1934: 133-134). Circa sechzig Jahre später bestätigt der Sprachatlas von Syrien für Palmyra nicht nur die Gebräuchlichkeit des b-Modifikators, sondern dokumentiert auch, dass ʿam- als der „Verbmodifikator Präsens“ fungiert (vgl. Behnstedt 1997: Bd. I, Karten 159-160). Ob nun diese beiden nicht ganz übereinstimmenden Befunde nahelegen, dass seit Cantineau (1934) der Einfluss des Hauptstadtdialektes (Koinéisierung) zum Tragen gekommen ist, oder ob in der früheren Untersuchung ein eventuell gebräuchliches ʿam- schlicht unberücksichtigt geblieben ist, kann nicht mehr geklärt werden. Gegen eine Nichtberücksichtigung spricht jedoch der Umstand, dass Cantineau mit der Erwähnung des aktiven Partizips (siehe obiges Zitat) der Frage nach den verschiedenen formalen Ausdrucksmitteln eines „présent proprement dit“ (s.o.) durchaus gezielt nachgeht und dabei sogar die hierfür wichtige Frage der Verbsemantik erwähnt (ebd. 1934: 134, 138-139). Ein paralleler präsentischer bzw. progressivischer Gebrauch von ʿam- mit einer komplementären Gruppe von Verben, wie wir sie in anderen Dialekten Syriens haben (vgl. Cowell 1964, McCarus 1976), dürfte ihm bei solcher Genauigkeit wohl kaum entgangen sein. 105 Vgl. Borg 1985: 101, Roth 1987: 21, Fn. 1; das p-Präfix hat in der 1. Person Sg. das Allomorph ma(n)- (vgl. Borg 1985: 101). 106 Vgl. Tsereteli 1970: 291, Fn. 1.

134

Melanie Hanitsch

Insgesamt fällt auf, dass sich solche „bi-only-Dialekte“ schwerpunktmäßig am östlichen Rand des b(i)-Verbreitungsgebietes in der Kontaktzone zwischen sesshaft dominierten und nomadisch dominierten Dialektzonen finden.107 Daher überrascht auch wenig, dass hier oftmals neben dem b(i)-Imperfekt auch ein grammatikalisiertes Sitzend-Imperfekt gebräuchlich ist, z.B. im Ḥōrān gāʿed yebči ‚il est en train de pleurer‘, gāʿed yemši ‚il est en train de marcher‘.108 Soukhne und seine zwei Ableger ič-Čōm/ǝl-Kōm und ǝṭ-Ṭayybi weisen sogar die stark reduzierten Formen kāʿ- und tāʿ- auf, z.B. kāʿ-tixbuz ‚sie ist beim Backen‘, tāʿ-yinkari ‚er wird (jetzt) gelesen‘.109 In den bereits genannten zentralpalästinischen Dialekten wird ein solcher Sitzend-Modifikator sogar mit dem b(i)- zu einem dritten Formtyp der doppelten Aspektmarkierung verbunden, nämlich zum /qāʿid bi-Imperfekt/.110 Diesen Formtyp kann man – in Anlehnung an die beiden bisher angesprochenen und am weitesten verbreiteten Formtypen – nicht nur formal als „hybrid“ bezeichnen (s.u. Abschnitt 4.), sondern auch, ebenso wie das „bi-only-Merkmal“, als ein Kontaktphänomen der sesshaft-nomadischen Übergangszone ansehen. Die „ʿam-only-Dialekte“, also Varietäten, von denen sicher gesagt werden kann, dass in ihnen nur Modifikatoren aus ʿammāl vorliegen, aber – von reinen Koine-Einflüssen einmal abgesehen – keinerlei b(i)-Modifikatoren, finden sich in den westlichen Oasendialekten Ägyptens, möglicherweise auch in etlichen 107 Zu Soukhne vgl. Behnstedt 1994: XVI-XVII, zum syrischen Raum insgesamt Behnstedt 1997, Bd. I, welchem zufolge die Dialekte der Oasen Palmyra/Tadmur und Soukhne den östlichsten Zipfel des weit in das Gebiet der Beduinendialekte (ebd. Karten 511-518) hineinragenden zentralsyrischen Dialektkontinuums („Steppendialekte und Dialekte am Steppenrand“) darstellen (vgl. Karten Nr. 503-510) und entsprechend mit den Beduinendialekten auch schon etliche phonologische und morphologische Merkmale teilen (vgl. Karten 508-510). 108 Unter Verweis auf Feghali 1928: 77 stellt Cantineau fest: „Cet emploi de gāʿed est à peu près inconnu des parlers libanais (Cantineau 1946: 222)“. 109 Behnstedt 1994: 60-61; die Etymologie kāʿ- < *kāʿid unterstreicht Behnstedt dadurch, dass er die Suḫni-Formen dem geographisch benachbarten, beduinischen ǧāʿid gegenüberstellt, z.B. ǧāʿid txábiz ‚sie ist beim Backen’ (ebd. 1994: 60; zur räumlichen und strukturellen Nähe zum Beduinischen s.o. Fn. 104). kāʿ- und tāʿ- stehen in allen drei Ortschaften in freier Variation. Die Umbildung kāʿ- > tāʿ- erklärt Behnstedt innersprachlich durch eine Analogie mit dem futurischen Modifikator ta- (< AA ḥattā) (vgl. Behnstedt 1994: 61). 110 Für Soukhne hingegen schließt Behnstedt eine solche Kumulation ausdrücklich aus (Behnstedt 1994: 61).

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

135

Dialekten des mittleren Niltals,111 sowie in einigen ländlichen und nomadischen Dialekten Tunesiens, insbesondere in der tunesischen Sahara.112 Im Verhältnis zu den „bi-only-Dialekten“ liegen sie also an der entgegengesetzten, nämlich west-südwestlichen Peripherie des Kernausbreitungsgebietes von b(i)- und ʿammāl.113 Die Oase il-Baḥariyya hat die Varianten imma ~ (i)mmi-, ima-, mi-, am- umu~ um-, z.B. ina mmi-niktib ‚ich bin am Schreiben‘, mi-ndawwar ‚wir suchen‘.114 Der Modifikator dient hier dem Ausdruck des „Präsens“. Zum Ausdruck von Habitualität, also dort, wo in anderen Dialekten das b(i)-Imperfekt mehr oder minder stark gebräuchlich bis obligatorisch ist, verwendet der Dialekt von il-Baḥariyya ediertem Textmaterial nach zu urteilen (Drop – Woidich 2007) in „konservativer Manier“ ausschließlich das einfache Imperfekt, z.B. gōmitna mi nnōm wu ngūmu nṣallu, wu niži niṭṭámminu ‚[Tageslauf] Unser Aufstehen, wir stehen auf und beten, dann frühstücken wir‘.115 Die Oase Daḫla hat ʿammā- und amā-, Farafra hat mi- und Ḫarga ʿamma-, immi-, am-. Diese Formen werden ausdrücklich zu den in Mittel- und Oberägypten vertretenen *ʿammāl-Ableitungen ʿamma-, amma-, ʿama-, ʿam-, ʿa-, maʿ-, und ma-116 in Beziehung gesetzt,117 zumal auch sonst die strukturellen Bezüge, die zwischen den westlichen Oasendialekten ins ägyptische Kernland hinein bestehen, vor allem das mittlere Niltal betreffen.118

111 Zum ägyptischen Raum, in dem die Ableitungen von ʿammāl die Formen ʿamma-, amma-, ʿama-, ʿam-, ʿa-, maʿ-, und ma- annehmen, vgl. Behnstedt – Woidich 1985: Bd. II, Karten 219-221. 112 Vgl. Agius – Harrak 1987: 172-173. 113 Dieses Verteilungsbild wird nur durch das Ḫātūni, den einzigen auf syrischem Boden alteingesessenen Qǝltu-Dialekt, gestört. Dieser gebraucht neben dem Sitzend-Modifikator qēʿid (s.o.) auch relativ häufig ʿam-, was jedoch von Shabo Talay als ausgeprägter Koinéeinfluss gewertet wird (vgl. Talay 1999: 182). 114 Drop – Woidich 2007: 74. 115 Drop – Woidich 2007: 148. 116 Behnstedt – Woidich 1985: Bd. II, Karte Nr. 221 („Verbmodifikator Präsens“). 117 Vgl. Woidich 1993: 346, Drop – Woidich 2007: 74, Fn. 63. 118 Vgl. Woidich 1993: 348, 355; inwieweit auch in den Dialektgebieten entlang des Nils, in denen ʿammāl-Ableitungen (und nicht etwa bi- oder ba-) den „Verbmodifikator Präsens“ stellen, das bi-Imperfekt ausschließlich Koinéinfluss geschuldet ist, oder ob es ggf. als Habitualis gebräuchlich ist, kann ich derzeit nicht beurteilen.

136

Melanie Hanitsch

3.1.2

Dialekte, die beide Morpheme aufweisen

Der Großteil der jeweiligen Verbreitungsgebiete von ʿam(māl) und b(i)- überschneiden sich, oder anders ausgedrückt: Fast alle Dialekte, die den einen Modifikator aufweisen, gebrauchen in der einen oder anderen Weise auch den anderen. Für dieses Gros der Dialekte ist, wie im Falle der Sitzend-Modifikatoren, zwischen zwei Haupttypen von TAM-Systemen zu unterscheiden: 1.) Dialekte, in denen ʿam(māl) und b(i)- niemals in Kontaktstellung vorkommen (Abschnitt 3.1.2.1.) und 2.) Systeme, in denen Kontaktstellung grundsätzlich möglich ist (Abschnitt 3.1.2.2.). 3.1.2.1 Kontaktstellung ausgeschlossen Für vereinzelte Lokaldialekte des syrolibanesischen Raumes findet sich der explizite Hinweis auf eine Nicht-Kumulierbarkeit von ʿammāl-Ableitungen mit dem b(i)-Imperfekt, z.B. Tripoli (Libanon)119 oder in-Nabk (Syrien).  120 Für den Dialekt der Çukurova (Südtürkei) gilt dies möglicherweise ebenfalls.121 Zumeist jedoch lassen Dialekte, die grundsätzlich beide Modifikatoren verwenden, auch in irgendeinem Umfang Kontaktstellung bzw. Kumulation zu. 3.1.2.2 Kontaktstellung möglich Innerhalb dieser Gruppe, die mit Abstand den größten Teil des ʿam- und biVerbreitungsgebietes ausmacht, müssen sowohl mit Bezug auf den Umfang als auch auf den funktionalen Status der doppelten Aspektmarkierung im Gesamtsystem zwei hauptsächliche Typen von TAM-Systemen unterschieden werden. Typ I wird in herausragender Weise vertreten durch das Kairenische: Das bImperfekt ist das gängige bzw. unmarkierte Ausdrucksmittel des aktuellen Präsens/Progressiv/imparfait concomitant, z.B. ʾaṣlaha dilwaʾti bitlabbis ilʿarūsa ‚sie kleidet nämlich jetzt gerade die Braut an‘, ḥaḍrítak bitbuṣṣi kida lē ‚warum schauen sie so?‘.122 Der Modifikator ʿammāl ist in semantisch-funktionaler Hinsicht eher schwach grammatikalisiert und nicht so sehr als Progressiv (‚am VERB-en sein‘, gerade VERB-en‘) aufzufassen. Vielmehr neigt er noch stark einem Kontinuativ zu (‚weiter-VERB-en‘) bzw. ist beim Ausdruck aktuell vorliegender

119 Vgl. El-Hajje 1954: 50. 120 Vgl. Gralla 2007: 126. 121 Die Paradigmen in Prochazka (2002: 113-114) lassen dies jedenfalls vermuten. 122 Woidich 2006: 281.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

137

Sachverhalte noch stark als Intensiv konnotiert (in etwa ‚immerzu VERB-en‘), z.B. ma-tʾaxzunīš, ʿammāla atkallim wi nsīt aṣabbaḥ ʿalēkum ‚nichts für ungut, ich rede in einer Tour und habe vergessen, euch ‚Guten Morgen‘ zu sagen‘, ʿammāl yibaḥlaʾ fīha ‚er glotzt sie immerzu an‘.123 Nicht-Kumulation von ʿammāl und bi-, wie in den angeführten Beispielen, ist der Regelfall. Kumulation ist der seltenere Fall, z.B. ʾāʿid gambi ʿAzīza w ʿammāl biybuṣṣilha ‚er saß neben ʿAzīza und schaute sie unverwandt an‘.124 Auch der Dialekt von Kfar ʿAbīda (Libanon) mit den Modifikatoren ʿan- ~ man- und b(i)-, sowie am südöstlichsten Rand des ʿam-Verbreitungsgebietes der Dialekt von Mekka (Saudiarabien) mit ʿammāl (+ Personalsuffix) können wohl diesem Typ zugeordnet werden.125 Typ II umfasst die flächenmäßig größere Gruppe von Dialekten, in denen das bloße b(i)-Imperfekt nicht mehr zum Ausdruck des aktuell vorliegenden dynamischen Sachverhalts (also als Progressiv) verwendet werden kann. Es wurde in dieser Funktion vollständig durch das ʿam (bi-)-Imperfekt abgelöst und nun somit auf seine nicht-episodischen Funktionen (Habitualis, etc.) und den futurischen Präsensgebrauch beschränkt. Hinsichtlich des Vorkommens der doppelten Aspektmarkierung /ʿam-b-Imperfekt/ gegenüber der gleichbedeutenden126 einfachen Aspektmarkierung /ʿam-Imperfekt/ liegt in Typ II ein Kontinuum vor, innerhalb dessen – vorbehaltlich weiterer Detailstudien – zunächst zwei regionale Schwerpunkte ausgemacht werden können: Der syrolibanesische Raum weist beim Ausdruck des aktuellen Präsens/Progressiv/ imparfait concomitant ein hohes Maß an innersprachlicher wie diatopischer Variation zwischen einfacher Aspektmarkierung (ʿam-Imperfekt) und doppelter Aspektmarkierung (ʿam-b-Imperfekt) auf.127 Im urbanen Palästinensischen hingegen ist ein starker Hang zur Kumulation zu beobachten.128 Im Folgenden werden in der Reihenfolge der steigenden Gebräuchlickeit der doppelten Aspektmarkierung einige Dialekte vorgestellt:

123 Woidich 2006: 283. 124 Ebd. 2006: 283. 125 Vgl. Feghali 1919: 138, Schreiber 1970: 44. 126 Vgl. Lentin 1994: 298. 127 Vgl. Feghali 1928: 40, Grotzfeld 1980: 182. 128 Für den allgemeinen Hinweis auf den starken Hang des Palästinensisch-Arabischen zur Kumulation von ʿam- und b(i)- danke ich Nora Boneh (Jerusalem).

138

Melanie Hanitsch

Das Damaszenische ist im Vergleich zu anderen Dialekten des syrischen Raumes der doppelten Aspektmarkierung eher abgeneigt. Am stärksten vertreten ist sie in der 1. Person Sg, in den übrigen Personen hingegen ist sie vergleichsweise selten anzutreffen.129 Unabhängig von der Person wird die Variation ʿam-Imperfekt ~ ʿam-b-Imperfekt als frei beschrieben.130 In Vertauschproben zeichnet sich aber ab, dass die doppelte Markierung insgesamt mit einer höheren Spezifizität assoziiert ist, z.B. mit einem spezifischeren Objekt: ʾāʿed ʿam-ʾǝsmaʿ musīʾa ‚ich (sitze hier und) höre gerade Musik’ à ʾāʿed ʿam-bǝsmaʿ la-ʿAbd ǝl-Ḥalīm Ḥāfeẓ ‚ich (sitze hier und) höre gerade ʿAbd al-Ḥalīm Ḥāfiẓ’ [Da-Ṭ].131 Im Dialekt von Baskinta (Libanon) ist die Präferenz der doppelten Aspektmarkierung für die 1. Person Sg. nicht dokumentiert. Im Großteil des Paradigmas ist es frei verwendbar. Nur in den Personen mit dem Imperfektpersonalpräfix ti- und ni- (in Abgrenzung zu t-K und n-K) ist die einfache Aspektmarkierung durch ʿan- weitgehend die Regel.132 Das urbane Palästinensische stellt mit Blick auf die Kumulation sozusagen die „Steigerung“ der soeben für Baskinta beschriebenen Situation dar. Auf ʿamfolgt obligatorisch b-, nur in den Personen mit dem Imperfektpräfix t(v)- oder n(v)- kann das b- ausfallen,133 vor t-K und n-K hingegen bleibt es stehen: ʿambakteb, ʿam-(b)tekteb, ʿam-(b~m)nekteb, ʿam-beketbu ‚ich schreibe, du (m.) schreibst, wir schreiben, sie schreiben‘, aber ausschließlich doppelte Markierung in ʿambaqīs, ʿam-betqīs, ʿam-menqīs, ʿam-beqīsu ‚ich messe, du (m.) misst, wir messen, sie messen‘.134 Eine weitere Steigerung der (weitgehend) obligatorischen Kumulation ist die Fusion der *ʿammāl-Ableitung mit dem b-Modifikator. Dieser Fall ist im 129 Vgl. Cowell 22005 (1964): 320. 130 Vgl. Grotzfeld 1964: 59, 1965: 87, Cowell 22005 (1964): 320 und Lentin 1994: 298. 131 Man beachte auch den Gebrauch des Sitzend-Partizips in Grammatikalisierungsstufe I-II, die zeigt, dass im Damaszenischen sogar eine schwach grammatikalisierte „dreifache“ Aspektmarkierung vorliegt, die zugleich – mit Blick auf die beteiligten Elemente – als „hybrid“ angesehen werden kann (s.u. Abschnitt 3 „Hybride“ Formtypen). 132 Vgl. Abu-Haidar 1979: 86. 133 Vgl. Halloun 1988: 25-26, 42 und Durand 1996: 134-136 für Jerusalem, Geva-Kleinberger 2004: 124 für Haifa. 134 Die Verbalformen basieren auf Halloun 1988: 42, 47 und wurden modifiziert unter Berücksichtigung einiger Angaben bei Durand 1996: 134, 136.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

139

Sprachatlas von Syrien für einige versprengte Ortschaften in Zentralsyrien verzeichnet, vor allem aber für die nordsyrischen Dialekte der Region Aleppo, wo sich neben ʿammāl b- auch ʿamb-, ʿab- und ab-, und sogar ein regelrechtes Portemanteau-Morphem ʿabb- findet.135 Ein ähnlicher Formenbestand liegt in den Sesshaftendialekten der nach Westen hin angrenzenden Region Antiochien (Türkei) vor, vgl. die Verbalformen ʿabyišrab, abyišrab und bayišrab für ‚er trinkt gerade‘.136 3.2 Diachronische Fragestellungen Im Vergleich zur doppelten Sitzend-Modifikation ist es bei der ʿam-biModifikation deutlich schwerer, aus der typologisch-dialektgeographischen Einbettung heraus diachronische Schlüsse zu ziehen. Die Diversität der zeitgenössischen Dialekte gewährt einen recht guten Einblick in die Zwischenstufen der Evolution der einfachen Sitzend-Modifikation und zwar hinsichtlich 1.) formaler Reduktion, 2.) semantischer Entwicklung (Stufe/Typ E-I bis III) und 3.) funktionaler Einbettung ins Gesamtsystem. Dadurch lässt sich auch die Ausbildung der doppelten/pleonastischen Aspektmarkierung /qāʿid Kv-/ (Stufe/Typ D-I bis III) selbst recht gut nachvollziehen und zu den Grammatikalisierungsgraden der jeweils beteiligten Modifikatoren in Bezug setzen. Im Falle der kumulativen Aspektmarkierung /ʿam-b-/ ist die Datenlage deutlich schwerer zu interpretieren. Im Folgenden sollen daher nur einige Gesichtspunkte herausgearbeitet werden, die bei künftigen, vertiefenden Untersuchungen zur Entstehung des doppelten ʿam bi-Imperfekts – sowie auch bei weiteren Untersuchungen zur Etymologie des bi-Modifikators selbst – berücksichtigt werden sollten: Die Modifikatoren ʿam(māl) und bi- sind, jeder für sich genommen, funktional stärker grammatikalisiert (Stufen II-III, keine Stufe I)  137 und in der Gesamt-

135 Vgl. Sabuni 1980: 110, 211-212, 213, 215, Lentin 1994: 298, Behnstedt 1997: Bd. I, Karte Nr. 161. 136 Vgl. Arnold 1998: 117. 137 Dies gilt zumindest für den Gebrauch von ʿammāl in modernen dialektalen Zeugnissen. In späten mittelarabischen Zeugnissen finden sich jedoch durchaus Belege, in denen ʿammāl im weitesten Sinne als Stufe I angesehen werden kann, insofern als seine lexikalische Bedeutung weitgehend bewahrt ist und es im Verhältnis zu den nachfolgenden Elementen der Phrase als Nukleus (und nicht als Modifikator)

140

Melanie Hanitsch

schau der Dialekte in viel größerem Umfang obligatorisiert als die vollen und reduzierten Sitzend-Modifikatoren. Auch die Kumulation von ʿam- und bi- ist im Neuarabischen deutlich stärker ausgeprägt als der Gebrauch des pleonastischen /qāʿid Kv-Imperfekts/: •

•

•

Im Dialektvergleich ist sie deutlich häufiger belegt, und das in einem weitgehend geschlossenen Gebiet von Unterägypten bis nach Nordsyrien und bis in die arabische Halbinsel hinein (Mekka). Sie deckt den Großteil des Verbreitungsgebietes eines jeden der beiden beteiligten Modifikatoren ab und lässt nur an der Peripherie kleinere Gebiete ohne Kumulation oder mit nur einem der beiden Modifikatoren stehen. Auch innersprachlich vermittelt das ʿam-bi-Imperfekt den Eindruck eines deutlich höheren Alters, insofern als es in zwei Regionen (urbanes Palästinensisch und Nordsyrien) einer Obligatorisierung als aktuelles Präsens/Progressiv/imparfait concomitant sehr nahe kommt und sogar fusioniert auftritt (Nordsyrien, Region Antiochien) – ein Phänomen, das für das Sitzend-Imperfekt nicht belegt ist.

Eine besondere Schwierigkeit in der diachronischen Deutung der typologischen Gliederung des ʿam- und bi-Gebietes betrifft Dialekte, in denen die Kumulation innersprachlich nur schwach ausgeprägt und auf Teile des Paradigmas beschränkt ist. So stellt sich etwa die Frage, ob die optionale Kumulation des Damaszenischen mit ihrer tendenziellen Beschränkung auf die phonologisch schwache 1. Person Sg. als Frühstufe des Eindringens des bi-Modifikators in das ʿam-Imperfekt Gefüge interpretiert werden muss, oder ob es vielmehr als Rest einer vormals ausgeprägteren, auf weitere Teile des Paradigmas ausgebreiteten und nunmehr erodierten Kumulation angesehen werden muss.138 Eine letzte, sicherlich gewagte Frage betrifft das Herz der Unterscheidung zwischen der pleonastischen Aspektmarkierung (doppelte Markierung durch Kognaten/Doubletten) und der kumulativen Aspektmarkierung (doppelte Aspektmarkierung durch nicht-kognate Morpheme): Stellt man die große heute bekannte Formenvielfalt, in der sich die Ableitungen aus *ʿammāl darbieten, einigen „extremen“ Ausprägungen des (mutmaßlichen) bi-Modifikators gegenüber, finden sich frappierende Bezüge. So ist der ʿammāl-Modifikator in der westlichen Peripherie bis hin zu mi- reduziert (s.o. il-Baḥariyya). Der im fungiert, z.B. ‫‚ ﻭوﺍاﻧﺎ ﻋﻣﺎﻝل ﻓﻲ ﺗﺷﺭرﻳﯾﺢ ﻋﻅظﺎﻣﻪﮫ‬ed io sono in solpare le sue [=gallina ] ossa (Lentin 1994: 300)‘. 138 Siehe auch die verschiedenen von Jérôme Lentin vorgeschlagenen Entwicklungsszenarien (vgl. Lentin 1994: 310).

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

141

gesamten Paradigma gleichlautende Modifikator m(i)- des Usbekistan-Arabischen wird hingegen im Allgemeinen mit dem ägypto-levantinischen b(i)-Imperfekt in einen etymologischen Bezug gebracht. 139 Einen Anhaltspunkt für diesen Bezug bietet wiederum das diatopisch weit verbreitete, wenn nicht gar im Dialektvergleich dominante m(i)- als Allomorph des b(i)-Imperfekts vor dem auf /n/ anlautenden Imperfektpräfix der 1. Person Pl. Gerade im westlichen Delta jedoch (also nicht allzuweit von den erwähnten, extrem reduzierten *ʿammāl-Ableitungen entfernt) findet sich dieses Allomorph des b(i)- gleich zweimal im Paradigma, da die betreffenden Dialekte ebenso wie die maghrebinischen in der 1. Person Sg. ebenfalls ein auf /n/ anlautendes Imperfektpräfix aufweisen.140 Im ebenfalls peripheren Dialekt von Kormatiki (Zypern) lautet das Allomorph der 1. Person Pl. sogar ma(n)-,141 was in bestechender Weise an die auf dem nahegelegenen Festland, nämlich im Libanon vertretenen *ʿammāl-Ableitung man-142 erinnert. Schließlich findet sich unter den a priori fusioniert erscheinenden Formen des nordsyrischen Raumes mindestens eine, die sich auch als eine schlichte Assimilation von ʿam- (also ohne Intervention von b(i)-) interpetieren lässt. So könnte etwa die Verbform ʿabtimši 143  ‚sie läuft‘ in diachronischer Hinsicht außer als *ʿa-btimši auch als *ʿab-timši (< ʿam-timši) mit Assimilation an das /t/ des Imperfektpräfixes analysiert werden.144 Betrachtet man diese vielfachen Bezüge zudem vor dem Hintergrund der in Abschnitt 3.1. deutlich gewordenen starken Überlappung der Verbreitungsgebiete beider Morpheme, ist die Versuchung groß, eine sehr alte und spätestens seit Kampff-

139 Vgl. Tsereteli 1970: 291, Fn. 1. 140 Es wird also im Allgemeinen von einer Assimilation des b(i)-Imperfekts an das /n/ des Imperfektpräfixes ausgegangen. 141 Vgl. Borg 1985: 101. 142 Vgl. Feghali 1919: 138-140 (Kfar ʿAbīda) und 1928: 39-40; Michel Féghali selbst leitet die Form man allerdings noch aus der Präposition min her (vgl. Feghali 1919: 138), Vgl. auch Cowell 22005 (1964): 320 (auch erwähnt in Agius – Harrak 1987: 169). Eine ähnliche Form ma- findet sich ebenfalls im Libanon und nach Norden hin anschließend entlang der syrischen Küste und in den alevitischen Dorfdialekten Antiochiens (Türkei) (vgl. Prochazka 2002: 113 unter Bezug auf Arnold 1998: 116). 143 Behnstedt 2000: 46. 144 In einigen syrischen Dialekten ist das b- Präfix genau diesem Kontext, also vor dem /t/ des Imperfektpräfixes, einer noch weiter reichenden Assimilation, nämlich Entsonorisierung unterworfen, z.B. in Palmyra (Tadmur) (vgl. Cantineau 1934: 132133) und im Ḥōrān (vgl. Cantineau 1946: 219).

142

Melanie Hanitsch

meyer (1900) verworfene145 Etymologie des bi-Modifikators einer erneuten Prüfung zu unterziehen, nämlich die von Vollers vorgeschlagene Herleitung des b(i)-Modifikators aus *ʿammāl: „Ich kann vielmehr in dem be des Präsens nichts anderes erkennen, als eine Umgestaltung des me der 1. Pers. Pl., welches seinerseits die letzte Verstümmelung des ʿammāl, ʿamma, ʿam, ma […] der gegenwärtigen Handlung ist“.146 Sollte sich erweisen, dass diese Etymologie im Lichte des heute verfügbaren Textmaterials eine zumindest ähnlich große Plausibilität hat wie die beiden heute dominanten Herleitungen aus bi- bzw. baynā (die ja jede für sich genommen ebenfalls ihre Schwachstellen haben),147 könnte die doppelte Aspektmarkierung durch ʿam-bi- leztlich ebenfalls als pleonastisch und damit als das Produkt eines reinforcements durch parallel gebräuchliche „Nebenformen“148 angesehen werden.149 4. „Hybride“ Formtypen Wie bereits in Abschnitt 3.1.1. angesprochen, kann der Sitzend-Modifikator, der in einigen „bi-only-Dialekten“ der östlichen Peripherie vorhanden ist, mit dem bi-Modifikator zu einer doppelten Aspektmarkierung verbunden werden. Dieser Fall ist insbesondere für die Dörfer um Ramallah, darunter auch Bīr Zēt 145 Vgl. Kampffmeyer 1900: 60-63. 146 Vollers 1887: 376; Karl Vollers zufolge fügt sich diese Etymologie bestens in die im Ägyptisch-Arabischen häufig zu beobachtenden Wechsel zwischen den bilabialen Lauten /b/ und /m/ ein, dessen herausragendster grammatikalisierter Repräsentant sicherlich matāʿ ‚Besitz‘ > bitāʿ Genitivexponent ist (nach ebd. 1887: 375-376). 147 Vgl. die kurze Zusammenfassung von Pro und Contra beider Etymologien in Rubin 2005: 150-151. 148 Vgl. Kampffmeyer (1900: 62) im weiteren Kontext seiner Auseinandersetzung mit Vollers Etymologie b- < *ʿammāl. 149 Es ist allerdings vorwegzunehmen, dass das wohl schwerwiegendste Argument gegen Vollers Etymologie ʿammāl > bi-, das im Übrigen bereits von Georg Kampffmeyer vorgebracht wurde, auch heute noch gilt: ʿammāl taucht erst sehr spät in mittelarabischen Texten auf, nämlich erst ab dem 17., vielleicht ab dem 16. Jh. (vgl. Kampffmeyer 1900: 62-63, Lentin 1994: 300). Dieser Umstand ist nur dann mit Vollers Etymologie in Einklang zu bringen, wenn man entweder annimmt, dass ʿammāl aufgrund einer starken Substandard-Konnotation keinen Eingang in die Texte gefunden hat („Ainsi l’absence d’une forme ne signifie-t-elle pas nécessairement son inexistence („censure“ stylistique)“ (Lentin 1994: 299)), oder aber, dass die zwei Grammatikalisierungswellen 1.) *ʿammāl > b(i)- und 2.) ʿammāl > ʿamma-, ʿam- zeitlich stark versetzt stattgefunden haben. Dies würde allerdings den Reiz und vielleicht auch die Plausibilität der Annahme dieser Etymologie schmälern.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

143

belegt,150 z.B. šū nti kāʿda bitsawwi ‚was tust du gerade?‘.151 Die Kumulation der Modifikatoren ist nicht obligatorisch, z.B. w-hī kāʿde tuẓrub ‚während sie gerade schlug‘.152 Doch Haim Blanc zufolge ist sie schon sehr verallgemeinert,  153 so dass eine starke Ähnlichkeit mit der sprachlichen Situation in Bagdad vorliegt. Schließlich soll nicht unerwähnt bleiben, dass sich auch im Zentrum der überlappenden ʿam- und b(i)-Verbreitungsgebiete, wo Kumulation grundsätzlich zulässig ist, eine ansatzweise Grammatikalisierung des aktiven Partizips ‚sitzend‘ abzeichnet. Wie schon die Vertauschprobe in Abschnitt 3.1.2.2. beiläufig illustrierte, ist etwa im Damaszenischen zu beobachten, dass eine Reihe von Verbalhandlungen, die typischerweise im Sitzen stattfinden, in beinahe idiomatischer Weise als Zustandsatz im ʿam b(i)-Imperfekt einem Partizip ʾāʿed ‚sitzend‘ untergeordnet werden, z.B. šu ʿam-taʿmel – ʾāʿed ʿam-bǝḥḍar fǝlǝm ‚was machst du gerade? – ich sitze (hier) und schaue einen Film (oder auch nur: ich schaue gerade einen Film)’ [Da-Ṭ]. Es scheint als würde hier eine „szenische Umstandsbeschreibung“ des Sitzens im Zusammenhang mit bestimmten Verben zunächst stark idiomatisiert, worauf diese Wendung dann ihrerseits eine (potentielle) Quellkonstruktion für die Grammatikalisierung einer doppelten bis dreifachen Aspektmarkierung bietet. 5. Sonstige Formtypen 5.1 Reinforcement weiterer Modifikatoren durch eine der behandelten Formen Analog zum palästinensischen /kēʿid bi-Imperfekt/ tritt der Sitzend-Modifikator in Marokko auch vor den Modifikator ka- (vermutlich < *kāʿin  oder kān154) 150 Vgl. Blanc 1960: 145-146, Seeger 2013: Bd. II, 208-209. 151 Seeger 2013: Bd. III, 208; besonders interessant ist, dass der Sitzend-Modifikator in den Dörfern um Ramallah auch dem bi-Imperfekt nachgestellt werden kann. Seegers Beispiele für den vorangestellten Konstruktionstyp sind m. E. allesamt Stufe II (mit sitzender Position kompatibel), während die nachgestellten allesamt Stufe III sind (mit sitzender Position inkompatibel). Inwiefern dieser erste Eindruck verallgemeinert werden kann, und wie sich im Neuarabischen insgesamt die syntaktische Variabilität (z.B. auch Distanzstellung des Modifikators zum Verb) zum Grad der semantischen Grammatikalisierung verhält, muss weiterführenden Untersuchungen vorbehalten bleiben. 152 Vgl. Seeger 2013: Bd. III, 208. 153 Vgl. Blanc 1960: 146. 154 Vgl. die zusammenfassende Diskussion verschiedener Etymologien in Ferrando 1996: 126-141.

144

Melanie Hanitsch

bzw. ta-, der ja für sich genommen in seiner Funktion stark dem bi-Modifikator des Ostens gleicht. Dabei begegnet die Quellsemantik des Sitzens allerdings in lexikalisch vielfältiger Form. So ist die Form gāʿǝd ‚assis‘ neben bārǝk ‚accroupi, assis, restant‘ und mrǝyyǝḥ ‚se reposant‘ auf ländliche Dialekte beschränkt,155 z.B. gāʿda kanṣǝbbǝn ‚Je suis occupée à laver le linge‘,156 bārǝk/mrǝyyǝḥ kaylʿǝb ‚Il est en train de jouer‘.157 In genuin urbanen Dialekten hingegen ist nur gālǝs ‚sitzend‘ anzutreffen, z.B. kān gālǝs ta-ylʿǝb hūwa u gālǝs tayṣūg-ha wāḥǝd-ǝs-sāʿa u hīya tǝwqǝf l-u… ‚Lui il était en train de jouer et il était en train de le conduire (un vélo), et tout d’un coup voilà qu’il s’arrète (le vélo)…158 Dabei beobachtet Caubet, dass /gālǝs ka-/ von einer vom Lande stammenden, aber in Fes lebenden Sprecherin nur in einer Weise verwendet wird, die meiner Stufe II entspricht (Handlung im Sitzen), z.B. gālsīn kayāklu ‚Ils sont en train de (assis à) manger‘, während der Sprachgebrauch von Z[akia] Iraqui-Sinaceur, die in Fes geboren ist, auch Stufe III (Handlung ohne Bezug zum Sitzen) aufweist, z.B. gālǝs tayžri ‚Il est en train de courir‘.159 Für das reinforcement durch die Form ʿammāl liegen mir keine Beispiele vor. Doch gibt es im Maghreb eine derivationelle Parallele in der Form ḫǝddām – ebenso wie ʿammāl ein Intensivnomen der Bedeutung ‚arbeitend, tätig sein‘ – z.B. āš ḫǝddām katdīr ‚Qu’est-ce que tu es en train de faire?‘.160 Dass für reinforcement durch bi- bzw. eine formal vollere Kognate hiervon kein gesichertes Beispiel angeführt werden kann, liegt insofern auf der Hand, als die Etymologie von bi- eben genau in Ermangelung derartiger Doubletten bisher nicht abschließend ermittelt werden konnte, sondern nur mit einiger Wahrscheinlichkeit entweder auf die Präposition bi- oder auf die Konjunktion baynā zurückgeführt werden konnte. Inwiefern andere, in ihrer Semantik vergleichbare Konjunktionen im Neuarabischen in Positionen anzutrefen sind, die nahelegen, dass sie gerade im Begriff sind, Teil einer doppelten Aspektmarkie155 Die Formen bārǝk und mrǝyyǝḥ scheinen zudem besonders auf die Küstenregion beschränkt zu sein (vgl. Caubet 1996: 90, 92). 156 Ebd. 1996: 93. 157 Ebd. 1996: 92. 158 Ebd. 1996: 93. 159 Ebd. 1996: 93. 160 Ebd. 1996: 94 (Hinsichtlich der Parallele zum ʿammāl-Modifikator des Ostens bezieht sich Dominique Caubet auf einen mündlichen Hinweis von David Cohen).

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

145

rung zu werden, ist eine offene Frage, deren Beantwortung detaillierte Korpusanalysen voraussetzen würde. Möglicherweise ist die Tendenz des Damaszenischen, die Konjunktion des Zustandssatzes w- hinter das Personalpronomen zu rücken und damit in Juxaposition zum finiten Verb des Zustands-satzes zu bringen,161 in diesem Lichte zu sehen, z.B. ʾana fī waḥde baʿrǝfa… bǝtnaʾʾeṭ wǝšša kǝllo / hīye w-ʿam tǝštǝġel binaʾʾeṭ wǝšša ‚ich… es gibt da eine [Frau], die ich kenne… die tropft aus dem ganzen Gesicht. Während sie arbeitet (hīye w-ʿam tǝštǝġel), tropft ihr ganzes Gesicht vor Schweiß‘ [Da-UW]. Der Zustandssatz als solcher bietet den narrativen Hintergrund (‚am Arbeiten sein‘) zu dem im Vordergrund/Fokus stehenden Sachverhalt (‚schwitzen‘) dar. Kennzeichnung von Hintergrundinformation ist ein klassisch imperfektivisches Konzept, so dass es nicht überraschen würde, wenn die Konjunktion, die einen entsprechenden Nebensatz einleitet, als Teil der Aspektmarkierung reanalysiert und schließlich dem Verb zugeschlagen würde. Sollte die Annahme, dass im Damaszenischen eine solche Reanalyse bereits angestoßen wurde, richtig sein, dürfte der Wechsel der syntaktischen Position als Prozess der Kondensierung im Verhältnis zu Verbalphrase angesehen werden: Die Konjunktion w- ist zunächst Teil einer syntaktischen Einheit von größerer Komplexität (Satzverknüpfung), wechselt jedoch ihre Position im Gesamtsyntagma, um fortan nur noch ein einzelnes Wort oder die Verbalgruppe eines Satzes zu modifizieren.162 5.2 Reinforcement der behandelten Modifikatoren durch weitere Formen Die vor allem im Maghreb beheimatete163 und vereinzelt auch im Jemen anzutreffende 164 deiktische Partikel rā + Personalsuffix (< *rah ‚sieh!’), 165 hat im Maghreb nicht nur diskursstrukturierende Funktionen, sondern auch sekundär entwickelte, mehr oder minder ausgeprägte aspektuelle Nebenbedeutungen.166

161 Vgl. Cowell 2005 (1964): 532. 162 Vgl. Lehmann 1995 (1982): 164. 163 Vgl. Cohen 1984: 281-282, Boucherit 2006: Bd. I, 64. 164 Vgl. Vanhove 1995: 243. 165 Vgl. Landberg 1909: 485-501 (erwähnt in Vanhove 1995: 248). 166 Vgl. Cohen 1984: 282, Boucherit 1987: 12-18; auch Fadoua Chaara klassifiziert in ihrer Studie zum Aspekt im Marokkanisch-Arabischen den Gebrauch der Partikel ra-(PP) vor dem Prädikat als „periphrastische Aspektkategorie“ – allerdings ohne dies in irgendeiner Weise zu ihrer an anderer Stelle sehr ausführlichen Beschreibung des Aspektsystems in Bezug zu setzen oder das zugegebenermaßen sehr

146

Melanie Hanitsch

Tritt sie unmittelbar vor ein Verb im Imperfekt, kommt sie funktional einem aktuellen Präsens/Progressiv/imparfait concomitant gleich, z.B. rāhi tǝšrǝb l ḥlīb duka ‚maintenant, elle est en train de boire du lait’ (Algers).167 Am östlichsten Rand des maghrebinischen Verbreitungsgebietes verbindet sich der Modifikator rā-PP mit dem vollen Sitzend-Modifikator zu einer doppelten Aspektmarkierung, z.B. rā-nā gāʿedīn naktebū ‚stiamo scrivendo’.168 Etwas weniger deutlich aspektuell ausgeprägt, sondern noch recht stark in seinen älteren deiktischfokusmarkierenden Funktionen verankert ist rā-PP, wenn es in formal ähnlicher Weise vor das marokkanische ka-Imperfekt tritt, z.B. huwa ka-yhdar ʿla ras-u u-ʿla mratu u-s-sulṭan ka-yḥsab-lu rah ka-y-hdar-lu ‚er sprach über sich selbst und seine Frau, während der Sultan dachte, daß er über soundso sprach‘.169 Im Jemen (südliches Hochland) findet sich ein Hinweis darauf, dass der dort verbreitete Verbalmodifikator ḏī- (seltener ḏā-), 170 der vermutlich auf ein deiktisches Element zurückzuführen ist171 und somit in seiner Quellsemantik dem maghrebinischen rā-PP nahesteht, möglicherweise ursprünglich mit dem bi-Präfix des angrenzenden zentralen und nördlichen Hochlandes 172 zu einer doppelten Aspektmarkierung *ḏī-bi- verbunden werden konnte. Jedenfalls hat der Modifikator ḏī- mancherorts in der 1. Person Sg. das kuriose Allomorph ḏīb-,173 z.B. ḏībaktib ‚ich schreibe gerade‘.174 Die Form dieses Allomorphs kann dadurch erklärt werden, dass das Element b- einer mutmaßlichen doppelten Aspektmarkierung nur in der 1. Person erhalten geblieben ist, da es hier in intervokalischer Position stand,  175 während es in anderen Personen geschwun-

schwer zu fassende Verhältnis zwischen Deixis/Fokusmarkierung und Aspekt zu thematisieren (Chaara 2003: 102); vgl. weiter die in Fn. 80 genannte Literatur. 167 Boucherit 1987: 14. 168 Panetta 1943: Bd. II, 251. 169 Colin 1957: 27 (Zitat und Übersetzung nach Chaara 2003: 105). 170 Vgl. Diem 1973: 45-63. 171 Zu den deiktischen Elementen ḏā- und ḏī- im Neuarabischen im Allgemeinen vgl. Fischer 1953: 57-66, 179-186. Zu den deiktischen Elementen im Jemen im Besonderen vgl. Behnstedt 1985: Bd. I, Karten 51-55. 172 Vgl. Fn. 96. 173 Vgl. Diem 1973: 56, 61. 174 Ebd. 1973: 56. 175 Persönliche Mitteilung durch Herrn Professor Werner Diem.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

147

den ist, z.B. *ḏīb(i)nuktub > ḏīnuktub ‚wir schreiben gerade‘,176 oder sich in diesem Kontext nie ganz durchsetzen konnte.177 Ein ganz ähnliches Spektrum an deiktischen, diskursstrukturierenden und aspektuellen Funktionen wie das maghrebinische rā-PP weist auch die jemenitische Partikel gad auf. Zum Ausdruck des aktuellen Präsens/Progressivs/imparfait concomitant wird dem bi-Imperfekt nahezu obligatorisch ein gad-PP vorangestellt. Analog zum libyschen /rā-PP gāʿed/ kann die ṣanʿānische Verbindung also auch als doppelte Progressivmarkierung /gad-PP bi-/ interpretiert werden, z.B. gadū biyurgud ḏalḥīn ‚er ist gerade am Einschlafen’ [Ṣa-N]. 6. Ausblick auf die doppelte Tempus- und Modusmarkierung Beispiele für eine doppelte oder gar dreifache Markierung, die in semantischer Hinsicht in der Übergangszone zwischen Modalität (volitiv) und Tempus (futurisch) angesiedelt ist, bietet Kampffmeier, z.B. ‫ ﺍاﻧﺎﺑﺎ ﺍاﻟﺠﺒﻞ ﺍاﻟﻴﯿﻮﻡم ﺑﺎ ﺑﻘﻨﺺ‬anaba-l-ǵebal alyåm bå bògnoṣ ‚je veux aller à la montagne chasser aujourd’hui‘.178 Da die Elemente bā und b- hier vermutlich Kognate sind, bezeichnet Kampffmeyer selbst diesen Fall als „pleonastisch“.179 Das in der Einleitung (Abschnitt 1.) angeführte Beispiel für doppelte Modusmarkierung (Optativ) im Muslimisch-Bagdadischen, ḫal-da-ngūl ‚let’s say’,180 ist nicht nur „kumulativ“, sondern auch das Produkt einer in besonderer Weise motivierten Verstärkung (reinforcement) des optativischen da-Modifikators: So nimmt Haim Blanc an, dass es die Homonymie dieses vergleichsweise wenig gebräuchlichen Morphems mit dem sehr verbreiteten präsentischen bzw. progressivischen da- ist, die in besonderer Weise nach einer Disambiguierung verlangt.181 Die Verstärkung durch das synonyme ḫal- kann also als ein Fall von grammatikalisierter Homonymenflucht angesehen werden.

176 Verbalform mit * in Anlehnung an die belegte Form ḏīnuktub bei Diem (1973: 56). 177 In systematischer Hinsicht weist dieser Fall eine gewisse Ähnlichkeit mit dem Damaszenischen auf, wo das ʿam-b-Imperfekt vor allem in der phonologisch schwachen 1. Person Sg. auftritt, was als Rest einer ehemals stärker auf das Gesamtparadigma erstreckten doppelten Aspektmarkierung interpretiert werden kann (s.o. Abschnitt 3.1.2.2.). 178 Kampffmeyer 1900: 57, Fn. 1. 179 Vgl. ebd. 1900: 57, Fn. 1. 180 Blanc 1964: 116. 181 Vgl. ebd. 1964: 116.

148

Melanie Hanitsch

Eine letzte, in semantischer wie formaler Hinsicht besondere Art der doppelten TAM-Markierung besteht schließlich in der schlichten Reduplikation des Modifikators, z.B. ġā-ġā-ddʿī līh ‚et vraiment de tout coeur elle invoqua Dieu!’ (Nomadendialekt der Hǝl-l-ġāba, Marokko).182 Diese ist hier vermutlich deshalb möglich, weil das Element ġā (vermutlich < *ġayr)183 in diesem Dialekt noch stark zwischen der etymologisch zugrundeliegenden, gradmarkierenden Funktion184 und den sekundären, temporal-aspektuellen Funktionen laviert – also kein „prototypischer“ TAM-Marker ist. Eine Reduplikation ikonischen Charakters ist die Markierung des Kontinuativs mittels Doppelung der etymologisch ungeklärten Partikel dǝl im JüdischBagdadischen, z.B. lǝsfīni mášǝt mášǝt mášǝt dǝl dǝl lǝma wáṣlǝt lǝlkūfa ‚The boat sailed on and on [lit.: the boat sailed, sailed, sailed] till it came to al-Kufa‘.185 Jedoch verlassen wir mit diesem letzten Beispiel strenggenommen bereits das Gebiet der doppelten TAM-Markierung und betreten den Boden einer TAMRealisierung, die verstärkt durch syntaktische und idiomatische Verfahren geleistet wird und deren syntagmatische Ausdehnung – wie eingangs erwähnt – zuweilen nur schwer abgesteckt werden kann.

Literaturverzeichnis Abu-Absi, Samir. 1995. Chadian Arabic. München – Newcastle: Lincom Europa (Languages of the World/Materials, 21). Abu-Haidar, Farida. 1979. A Study of the Spoken Arabic of Baskinta. Leiden [u.a.]: Brill. –––– 1991. Christian Arabic of Baghdad. Wiesbaden: Harrassowitz (Semitica Viva, 7). –––– 1994. „Les particules préverbales dans le dialecte musulman de Bagdad“, in: Dominique Caubet – Martine Vanhove (Hrsg.). Actes des premières journées internationales de dialectologie arabe de Paris. Paris: Institut national des langues et civilisations orientales (INALCO), S. 151-160. 182 Guironnet 1954: 429. 183 Vgl. ebd. 1954: 425; man vergleiche auch die in der Einleitung (Abschnitt 1.) erwähnte Partikel ġar(r) des Ṣanʿānischen. 184 Vgl. ebd. 1954: 427-428. 185 Mansour 1985: 66/67; vgl. auch die ikonisch gelängte Variante des (vermutlich) selben Elements im Muslimisch-Bagdadischen, z.B. ʾawwal bārḥa riḥna_ṣ-ṣōb_il-karḫ u-ẓall it-taksi yimši bīna dīīī ʾila ʾan wiṣalna (Fakhry 2015: keine Seitenzahl) ‚Vorgestern sind wir mit dem Taxi in den Westteil Bagdads (il-karḫ) gefahren. Das Taxi fuhr und fuhr… bis wir (schließlich) ankamen (Übersetzung in Anlehnung an ebd. 2015: keine Seitenzahl)‘.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

149

Agius, Dionisius A. – Amir Harrak 1987. „Auxiliary particles preceding the imperfective aspect in Arabic dialects”, in: Arabica, Revue d’Etudes Arabes 34: 164-180. Arnold, Werner. 1998. Die arabischen Dialekte Antiochiens. Wiesbaden: Harrassowitz (Semitica Viva, 19). Bauer, Leonhard. 41926 (1872). Das palästinische Arabisch. Die Dialekte des Städters und des Fellachen. Grammatik, Übungen und Chrestomathie. Leipzig: J. C. Hinrich’sche Buchhandlung. Behnstedt, Peter. 1985. Die nordjemenitischen Dialekte. Teil 1. Atlas. Wiesbaden: Reichert (Jemen-Studien, Bd. 3). –––– 1987. Die Dialekte der Gegend von Ṣaʿdah (Nord-Jemen). Wiesbaden: Harrassowitz (Semitica Viva, 1). –––– 1994. Der arabische Dialekt von Soukhne (Syrien). Teil 2. Phonologie, Morphologie, Syntax. Teil 3. Glossar. Wiesbaden: Harrassowitz (Semitica Viva, 15, 2-3). –––– 1997. Sprachatlas von Syrien. I, 1. Kartenband. Wiesbaden: Harrassowitz (Semitica Viva, 17). –––– 2000. Sprachatlas von Syrien. II. Volkskundliche Texte. Wiesbaden: Harrassowitz (Semitica Viva, 17). Behnstedt, Peter – Manfred Woidich. 1985. Die ägyptisch-arabischen Dialekte. II: Dialektatlas von Ägypten. Wiesbaden: Reichert. Blanc, Haim. 1964. Communal Dialects in Baghdad. Cambridge (Massachusetts): Harvard University Press (Harvard Middle Eastern Monographs, 10). Blau, Joshua. 1960. Syntax des palästinensischen Bauerndialekts von Bīr Zēt auf Grund der „Volkserzählungen aus Palästina“ von Hans Schmidt und Paul Kahle. Walldorf-Hessen: Verlag für Orientkunde Dr. H. Vorndran (Beiträge zur Sprach- und Kulturgeschichte des Orients, 13). Borg, Alexander. 1985. Cypriot Arabic. A historical and comparative investigation into the phonology and morphology of the Arabic vernacular spoken by the Maronites of Kormakiti village in the Kyrenia district of North-Western Cyprus. Wiesbaden – Stuttgart: Steiner (Abhandlungen für die Kunde des Morgenlandes, 47-4). Boucherit, Aziza. 1988. „Note sur l’expression de la concomitance dans le système verbal de l’arabe parlé à Alger” in: Matériaux Arabes et Sudarabiques. Recherches en cours 1987. Paris.: Geuthner – Association GELLAS (Université Paris III), S. 11-20. –––– 2006. „Algiers Arabic” in: Cornelis H. M. Versteegh [u.a.] (Hrsg.). Encyclopedia of Arabic Language and Linguistics. I: A-Ed. Leiden [u.a.]: Brill, S. 58-66. Brustad, Kristen. 2000. The Syntax of Spoken Arabic. A comprehensive study of Moroccan. Egyptian, Syrian, and Kuwaiti Dialects. Washington D. C.: Georgetown UP. Bussmann, Hadumod (Hrsg.). 32002 (1983). Lexikon der Sprachwissenschaft. Stuttgart: Alfred Kröner Verlag. Bybee, Joan – Revere Perkins – William Pagliuca. 1994. The Evolution of Grammar. Tense, aspect, and modality in the languages of the world. Chicago – London: The University of Chicago Press.

150

Melanie Hanitsch

al-Čalabī , Dāwūd. 1935 (1354). al-ʾĀṯār al-ʾārāmīya fī luġat al-Mawṣil alʿāmmīya. al-Mawṣil: Maṭbaʿat an-naǧm al-kaldānīya. Cantineau, Jean. 1934. Le Dialecte Arabe de Palmyre. Tome I. Grammaire. Beirut: Institut Français de Damas. –––– 1938. „Remarques sur les parlers de sédentaires syro-libano-palestiniens”, in: Bulletin de la Société de Linguistique de Paris 40.1: 80-88. –––– 1946. Les parlers Arabes du Ḥōrān. Notions générales, grammaire. Paris: Librairie C. Klincksieck (Collection linguistique de la Société de Linguistique de Paris, 52). Campbell, Lyle. 2001. „What’s wrong with grammaticalization?”, in: Lyle Campbell (Hrsg.). Grammaticalization: a critical assessment. Oxford [u.a.]: Pergamon Press, S. 113-161 (Language Sciences 23,1-2 (Sonderband)). Caubet, Dominique. 1996. „gāles kayxdem, xāyǝḍ kayxdǝm: Approche sociolinguistique de l’expression de la concomitance en Arabe Marocain“, in: Estudios de dialectología norteafricana y andalusí 1: 87-100. Chaara, Fadoua. 2003. Die Interaktion der Aspektsemantik mit dem Lexikon im Marokkanisch-Arabischen. München: Lincom Europa (zugl. Köln, Univ., Diss., 2002). Cohen, David. 1975. Le parler arabe des juifs de Tunis. II: Étude linguistique. Paris: Mouton. –––– 1984: La phrase nominale et l’évolution du système verbal en sémitique. Etudes de syntaxe historique. Leuven – Paris: Editions Peeters (Collection linguistique publiée par la Société de Linguistique de Paris, 72). Cohen, Marcel. 1924. Le système verbal sémitique et l’expression du temps. Paris: Leroux (Publications de l’Ecole des Langues Orientales Vivantes, Ve série, 11). Colin, Georges S. 1957. Recueil de textes en arabe marocain. Contes et anecdotes. Paris: Maisonneuve. Cowell, Mark W. 22005 (1964). A Reference Grammar of Syrian Arabic. Based on the Dialect of Damascus. Washington D.C.: Georgetown University Press. Czapkiewicz, Andrzej. 1986. „The new indicative in Modern Arabic dialects“, in: Collectanea Linguistica 53: 43-47. Denz, Adolf. 1971. Die Verbalsyntax des neuarabischen Dialektes von Kwayriš (Irak). mit einer einleitenden allgemeinen Tempus- und Aspektlehre. Wiesbaden: Steiner. Diem, Werner. 1973. Skizzen jemenitischer Dialekte. Beirut: F. Steiner (Beiruter Texte und Studien, 13). Wiesbaden: Steiner. Drop, Hanke – Manfred Woidich. 2007. ilBaḥariyya. Grammatik und Texte. Wiesbaden: Harrassowitz (Semitica Viva, 39, Ägyptische Dialekte, Teil I). Durand, Olivier. 1996. Grammatica di Arabo Palestinese. Il dialetto di Gerusalemme. Prefazione di Wasim Dahmash. Roma: Università degli Studi „La Sapienza“. Eisele, John C. 1999. Arabic Verbs in Time. Tense and aspect in Cairene Arabic. Wiesbaden: Harrassowitz (Semitica Viva, 20). Eksell, Kerstin. 1995a. „Some punctual and durative auxiliaries in Syro-Palestinian dialects”, in: [Joe Cremona – Clive Holes – Geoffrey Khan (Hrsg.)]. Proceedings of the 2nd International Conference of L’Association Internationale pour la Dialectologie Arabe

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

151

held at Trinity Hall in the University of Cambridge 10-14 September 1995. [Cambridge: University Publications Centre – Middle East Centre], S. 41-49. –––– 1995b. „Complexity of linguistic change as reflected in Arabic dialects”, in: Dialectologia Arabica. A Collection of Articles in Honour of the Sixtieth Birthday of Professor Heikki Palva. Helsinki: Finnish Oriental Society, S. 63-73 (Studia Orientalia, 75). –––– 1998. „On some verbal modificators in Neo-Aramaic. With a short comparison to Arabic dialects”, in: Acta Orientalia 59: 52-74. –––– 2005. „The origin and development of the cursive b-imperfect in Syrian Arabic”, in: Kerstin Eksell – Thora Vinther (Hrsg.). Change in Verbal Systems. Issues on Explanation. Frankfurt a. M. [u.a.]: Peter Lang, S. 73-98. El-Hajjé, Hassan. 1954. Le parler Arabe de Tripoli (Liban). Paris: Klincksieck. Erwin, Wallace. M. 1963. A Short Reference Grammar of Iraqi Arabic. Washington D.C.: Georgetown University Press. Fakhry, Salah A. [voraussichtlich 2015]. „Tempus, Aspekt und Modalität im BagdadArabischen”, in: Lutz Edzard (Hrsg.). The Morpho-Syntactic Encoding of Tense and Aspect in Semitic. Wiesbaden: Harrassowitz. Feghali, Michel T. 1919. Le parler de Kfarʿabīda (Liban-Syrie). Essai linguistique sur la phonétique et la morphologie d’un parler Arabe Moderne. Paris: Editions Ernest Leroux. –––– 1928. Syntaxe des parlers Arabes actuels du Liban. Paris: Librairie Orientaliste Paul Geuthner (Bibliothèque de l’Ecole des Langues Vivantes, 9). Ferrando, Ignacio. 1996. „Quelques observations sur l’origine, les valeurs et les emplois du préverbe ka- dans les dialects arabes occidentaux (maghrébins et andalous)”, in: Matériaux Arabes et Sudarabiques. MAS-GELLAS 1995-96. Paris: GELLAS (Université Paris III), S. 115-143. Fischer, Wolfdietrich 1953. Die demonstrativen Bildungen der neuarabischen Dialekte. Ein Beitrag zur historischen Grammatik des Arabischen (Erlangen, Univ., Diss. 1953). Fischer, Wolfdietrich – Otto Jastrow (Hrsg.). 1980. Handbuch der arabischen Dialekte. Mit Beiträgen von P. Behnstedt, H. Grotzfeld, B. Ingham, A. Sabuni, P. Schabert, H.-R. Singer, L. Tsotskhadze und M. Woidich. Wiesbaden: Otto Harrassowitz (Porta Linguarum Orientalium, Neue Serie, 16). Geva-Kleinberger, Aharon. 2004. Die arabischen Stadtdialekte von Haifa in der ersten Hälfte des 20. Jahrhunderts. Wiesbaden: Harrassowitz (Semitica Viva, 29). Gralla, Sabine. 2007. Der arabische Dialekt von Nabk (Syrien). Wiesbaden: Harrassowitz (Semitica Viva, 37). Grotzfeld, Heinz. 1964. Laut- und Formenlehre des Damaszenisch-Arabischen. Wiesbaden 1964. Wiesbaden: Kommissionsverlag Franz Steiner (Abhandlungen für die Kunde des Morgenlandes, 35, 3). –––– 1965. Syrisch-arabische Grammatik. Wiesbaden: Harrassowitz (Porta Linguarum Orientalium, Neue Serie, 8). –––– 1980. „Das syrisch-palästinensische Arabisch“, in: Wolfdietrich Fischer – Otto Jastrow (Hrsg.). Handbuch der arabischen Dialekte. Wiesbaden: Harrassowitz, S. 174206.

152

Melanie Hanitsch

Guironnet, Max. 1954. „ġā-, un outil linguistique polyvalent chez les Ūlād Yăʿgōb des Srāġna”, in: Hespéris: Archives berbères et bulletin de l’Institut des Hautes Études Marocaines 41, 425-433. Halloun, Moin. 1988. Tables for Paradigms. The complete verbal system of conjugation in the spoken Arabic of Jerusalem and the Galilee with an Arabic-English dictionary of 1000 English verbs. Bethlehem: Bethlehem University. Harrell, Richard S[lade]. 1962. A Short Reference Grammar of Moroccan Arabic. With an Appendix of Texts in Urban Moroccan Arabic by Louis Brunot. Washington D.C.: Georgetown University Press. Holes, Clive. 2001. Dialect, culture, and Society in Eastern Arabia. I: Glossary. Leiden [u.a.]: Brill (Handbook of Oriental Studies, Section One: The Near and Middle East, 51). –––– 2004. Structures, Functions and Varieties (revised edition). Washington, DC: Georgetown University Press. Hopper, Paul J. 1987. „Emergent Grammar“, in: Jon Aske [u.a.] (Hrsg.). Berkeley Linguistics Society. Proceedings of the Thirteenth Annual Meeting February 14-16, 1987. General Session and Parasession on Grammar and Cognition. Berkeley: Berkeley Linguistics Society, S. 139-157. Hopper, Paul J. – Elizabeth Closs Traugott 22003 (1993). Grammaticalization. Cambridge: Cambridge University Press. Jastrow, Otto. 1978. Die mesopotamisch-arabischen Qǝltu-Dialekte. I. Phonologie, Morphologie. Wiesbaden: Steiner (Abhandlungen für die Kunde des Morgenlandes, 46, 1). –––– 1979. „Zur arabischen Mundart von Mossul“, in: Zeitschrift für arabische Linguistik 2: 36-75. –––– 1980. „Das mesopotamische Arabisch”, in: Wolfdietrich Fischer – Otto Jastrow (Hrsg.). Handbuch der arabischen Dialekte. Wiesbaden: Harrassowitz, S. 140-173. –––– 1982. „Die Struktur des Neuarabischen“, in: Wolfdietrich Fischer – Helmut Gätje (Hrsg.). Grundriß der arabischen Philologie. Bd. I: Sprachwissenschaft. Wiesbaden: Reichert, 128-141. –––– 1990. Der arabische Dialekt der Juden von ʿAqra und Arbīl. Wiesbaden: Harrassowitz (Semitica Viva, 5). –––– 2004. „Jüdisches, christliches und muslimisches Arabisch in Mossul”, in: Martine Haak – Rudolf de Jong – Kees Versteegh (Hrsg.). Approaches to Arabic Dialects. A Collection of Articles presented to Manfred Woidich on the Occasion of his Sixtieth Birthday. Leiden – Boston: Brill, S. 135-150. –––– 2007. „Iraq“, in: Cornelis H. M. Versteegh [u.a.] (Hrsg.). Encyclopedia of Arabic Language and Linguistics. II. Eg – Lan, S. 414-424. Johnstone, T[homas] M. 1967. Eastern Arabian Dialect Studies. London – New York – Toronto: Oxford University Press (London Oriental Series, 17). De Jong, Rudolf. 2000. A Grammar of the Bedouin Dialects of the Northern Sinai Littoral. Bridging the linguistic gap between the Eastern and Western Arab World. Leiden – Boston – Köln: Brill (Handbuch der Orientalistik, Erste Abteilung: Der Nahe und Mittlere Osten, 52).

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

153

Kampffmeyer, G[eorg]. 1900. „Die arabische Verbalpartikel b (m). Beiträge zur Dialectologie des Arabischen II“, in: Mittheilungen des Seminars für Orientalische Sprachen an der Königlichen Friedrich Wilhelms-Universität zu Berlin 3, S. 48-101. Kimary, Shahin 22000 (1995). Rural Palestinian Arabic (Abu Shusha dialect). München: Lincom Europa (Languages of the World/Materials, 28). Kuryłowicz, Jerzy 1975. „The evolution of grammatical categories”, in: Jerzy Kuryłowicz. Esquisses linguistiques II. München: Wilhelm Fink Verlag, S. 38-54 (Internationale Bibliothek für Allgemeine Linguistik, 37). Landberg, Carlo 1909. Etudes sur les dialectes de l'Arabie Méridionale. Daṯînah. Deuxième partie: Commentaire des textes prosaiques. Leiden: Brill. Lehmann, Christian. 1995 (1982). Thoughts on Grammaticalization. München – Newcastle: Lincom Europa (Lincom Studies in Theoretical Linguistics, 1). Lentin, Jérôme. 1994. „ʿam yǝktob, ʿam byǝktob: Elements pour une approche historique“, in: Dominique Caubet – Martine Vanhove (Hrsg). Actes des Premières journées internationales de dialectologie arabe de Paris. Paris: Institut national des langues et civilisations orientales (INALCO), 1994, S. 297-316. Madouni, Jihane. 1983. „Les valeurs de la particule rā dans un parler de Sidi-BelAbbès”, in: Matériaux Arabes et Sudarabiques. Recherches en cours 1993 (Nouvelle Série) 5, Paris: Geuthner – Association GELLAS (Université Paris III), S. 123-136. Malaika, Nisar. 21963 (1959). Grundzüge der Grammatik des arabischen Dialektes von Bagdad. Wiesbaden: Harrassowitz. Mansour, Jacob. 1985. „The particle baqa in the Judaeo-Arabic Dialect of Baghdad”, in: Zeitschrift für arabische Linguistik 14: 62-75. –––– 1991. The Jewish Baghdadi dialect. Studies and texts in the Judaeo-Arabic Dialect of Baghdad. Or-Yehuda: Babylonian Jewry Heritage Center – The Institute for Research on Iraqi Jewry. McCarus, Ernest. 1976. „A semantic analysis of Arabic Verbs”, in: Louis L. Orlin [u.a.]. Michigan Oriental Studies in honor of George G. Cameron offered by the Faculty of the Department of Near Eastern Studies, The University of Michigan. Ann Arbor: Department of Near Eastern Studies (The University of Michigan), S. 3-28. Nöldeke, Theodor. 1904. „Über einige arabische Verbalpräfixe“, in: Theodor Nöldeke. Beiträge zur semitischen Sprachwissenschaft. Straßburg: Trübner, S. 63-68. Owens, Jonathan. 1984. A Short Reference Grammar of Eastern Libyan Arabic. Wiesbaden: Harrassowitz. ––––– 1993. A Grammar of Nigerian Arabic. Wiesbaden: Harrassowitz (Semitica Viva,10). Panetta, Ester 1943. L’Arabo parlato a Bengasi. II. Grammatica. Roma: La Libreria dello Stato. Persson, Maria. 2008. „The role of the b-prefix in Gulf Arabic dialects as a marker of future, intent and/or irrealis”, in: Journal of Arabic and Islamic Studies 8: 26-52. –––– 2009. „Circumstantial Qualifiers in Gulf Arabic Dialects“, in: Bo Isaksson – Heléne Kammensjö – Maria Persson. Circumstantial Qualifiers in Semitic. The case of Arabic

154

Melanie Hanitsch and Hebrew. Edited by Bo Isaksson. Wiesbaden: Harrassowitz, S. 206-289 (Abhandlungen für die Kunde des Morgenlandes, 70).

Piamenta, Moshe. 1991. Dictionary of post-classical Yemeni Arabic. II. A – Š. Leiden [u.a.]: Brill. Prochazka, Stephan. 2002. Die arabischen Dialekte der Çukurova (Südtürkei). Wiesbaden: Harrassowitz (Semitica Viva, 27). Reichmuth, Stefan. 1983. Der arabische Dialekt der Šukriyya im Ostsudan. Hildesheim [u.a.]: Olms. Retsö, Jan. 1982-1983. „Subjectless sentences in Arabic dialects“, in: Orientalia Suecana 31-32: 71-91. –––– 1983. The Finite Passive Voice in Arabic Dialects. Göteborg: University of Göteborg. –––– 1989. Diathesis in the Semitic Languages. A comparative morphological study. Leiden [u.a.]: Brill. –––– 2004. „Relative clause marking in Arabic Dialects”, in: Manfred Woidich [u.a.]. Approaches to Arabic Dialects. A collection of articles presented to Manfred Woidich on the occasion of his sixtieth birthday. Leiden – Boston: Brill, S. 263-273. –––– 2005. „The Number-Gender-Mood Markers of the Prefix Conjugation in Arabic Dialects”, in: Lutz Edzard – Jan Retsö (Hrsg.). Current Issues in the Analysis of Semitic Grammar and Lexicon I. Oslo-Göteborg Cooperation 3rd-5th June 2004. Wiesbaden: Harrassowitz, S. 31-40. –––– 2006. „Thoughts about the Diversity of Arabic”, in: Current Issues in the Analysis of Semitic Grammar and Lexicon II. Oslo-Göteborg Cooperation 3rd-5th June 2004. Wiesbaden: Harrassowitz, S. 23-33. –––– 2009. „The m-suffix with the finite verb 3rd and 2nd person plural in Arabic dialects”, in: Zeitschrift für arabische Linguistik 50: 39-50. –––– 2014. „The b-imperfect once again: typological and diachronic perspectives”, in: Lutz Edzard – John Huehnergard (Hrsg.). Proceedings of the Oslo–Austin Workshop in Semitic Linguistics, Oslo, May 23 and 24, 2013. Wiesbaden: Harrassowitz, S. 64-72. Rosenhouse, Judith. 1984. The Bedouin Arabic Dialects. General problems and a close analysis of North Israel Bedouin Dialects. Wiesbaden: Harrassowitz. –––– 2006. „Bedouin Arabic“, in: Cornelis H. M. Versteegh [u.a.] (Hrsg.). Encyclopedia of Arabic Language and Linguistics. I: A-Ed, S. 259-269. Rossi, Ettore. 1939. L’arabo parlato a Ṣanʿāʾ. Grammatica, testi, lessico. Rom: Istituto per l’Oriente. Roth, Arlette. 1987. „Questions à propos de quelques usages „particuliers“ de l’inaccompli à préfixe p- dans le parler arabe de Kormatiki (Chypre)”, in: Matériaux Arabes et Sudarabiques. Recherches en cours 1987. Paris: Geuthner – Association GELLAS (Université Paris III), S. 21-33. Rubin, Aaron D. 2005. Studies in Semitic Grammaticalization. Winona Lake (Indiana): Eisenbrauns.

„Doppelte“ Tempus- und Aspektmarkierung im Neuarabischen

155

Sabuni, Abdulghafur. 1980. Laut- und Formenlehre des arabischen Dialekts von Aleppo. Frankfurt a. M. – Bern – Cirencester (UK): Verlag Peter D. Lang (Heidelberger Orientalistische Studien, 2). Seeger, Ulrich 2013. Der arabische Dialekt der Dörfer um Ramallah. Teil 3: Grammatik. Wiesbaden: Harrassowitz (Semitica Viva, 44,3). Simeone-Senelle, Marie-Claude. 1986. „Systèmes aspecto-temporels en arabe maghrébin: Introduction.“, in: Matériaux Arabes et Sudarabiques. Recherches en cours 1985-86. Paris: Geuthner – Association GELLAS (Université Paris III), S. 57-61. Singer, Hans-Rudolf. 1980. „Das Westarabische oder Maghribinische“, in: Wolfdietrich Fischer – Otto Jastrow (Hrsg.). Handbuch der arabischen Dialekte. Wiesbaden: Harrassowitz, S. 249-265. –––– 1984. Grammatik der arabischen Mundart der Medina von Tunis. Berlin [u.a.]: de Gruyter. Schreiber, Giselher. 1970. Der arabische Dialekt von Mekka. Abriß der Grammatik mit Texten und Glossar (Münster, Univ., Diss. 1970). Socin, Albert. 1882–1883. „Der arabische Dialekt von Mosul und Märdin“, in: Zeitschrift der Deutschen Morgenländischen Gesellschaft 36: 1-53, 238-277. Spitta, Wilhelm. 1880. Grammatik des arabischen Vulgärdialectes von Aegypten. Leipzig: J. C. Hinrich’sche Buchhandlung. Stewart, Devin. 1998. „Clitic Reduction in the Formation of Modal Prefixes in the PostClassical Arabic Dialects and Classical Arabic Sa-/Sawfa”, in: Arabica. Journal of Arabic and Islamic Studies 45: 104-128. al-Tajir, Mahdi Abdalla. 1982. Language and Linguistic Origins in Bahrain. The Baharnah Dialect of Arabic. London: Kegan Paul. Talay, Shabo 1999. Der arabische Dialekt der Khawētna. I. Grammatik. Wiesbaden: Harrassowitz (Semitica Viva, 21,1). –––– 2011. „Arabic Dialects of Mesopotamia“, in: Stefan Weninger [u.a.] (Hrsg.). The Semitic Languages. An International Handbook. Berlin [u.a.]: Mouton – de Gruyter, S. 909-920 (Handbücher zur Sprach- und Kommunikationswissenschaft, 36). Tsereteli, George V. 1970. „The verbal particle m/mi in Bukhara Arabic“, in: Folia Orientalia 12: 291-295. Vanhove, Martine. 1993. La langue Maltaise. Etudes syntaxiques d’un dialecte arabe “périphérique”. Wiesbaden: Harrassowitz (Semitica Viva, 11). –––– 1995a. „Les particules qad et raʾ dans un dialecte arabe de la region de Yāfiʿ (Yémen)”, in: [Joe Cremona – Clive Holes – Geoffrey Khan (Hrsg.)]. Proceedings of the 2nd International Conference of L’Association internationale pour la Dialectologie Arabe held at Trinity Hall in the University of Cambridge 10-14 September 1995. [Cambridge: University Publications Centre – Middle East Centre], S. 243-252. –––– 1995b. „A propos du verbe dans les dialectes arabes de Yāfiʿ (Yémen)“, in: Dialectologia Arabica. A Collection of Articles in Honour of the Sixtieth Birthday of Professor Heikki Palva. Helsinki: Finnish Oriental Society, S. 257-269 (Studia Orientalia, 75).

156

Melanie Hanitsch

Vollers, Karl. 1887. „Beiträge zur Kenntnis der lebenden arabischen Sprache in Ägypten“, in: Zeitschrift der Deutschen Morgenländischen Gesellschaft 41: 365-402. Watson, Janet C. E. 1993. A Syntax of Ṣanʿānī Arabic. Wiesbaden: Harrassowitz (Semitica Viva, 13). Watson, Janet C. E. with ʿAbd al-Salām al-ʿAmri. 2000. Waṣf Ṣanʿā: Texts in Ṣanʿānī Arabic. Wiesbaden: Harrassowitz (Semitica Viva, 23). Wild, Stefan. 1964. „Die resultative Funktion des aktiven Partizips in den syrischpalästinischen Dialekten des Arabischen“, in: Zeitschrift der Deutschen Morgenländischen Gesellschaft 114: 239-254. Woidich, Manfred. 1980. „Das Ägyptisch-Arabische“, in: Wolfdietrich Fischer – Otto Jastrow (Hrsg.): Handbuch der arabischen Dialekte. Wiesbaden: Harrassowitz, S. 207229. –––– 1993. „Die Dialekte der ägyptischen Oasen: westliches oder östliches Arabisch?“, in: Zeitschrift für arabische Linguistik 25: 340-359. –––– 1995. „Some cases of grammaticalisation in Egyptian Arabic”, in: [Joe Cremona – Clive Holes – Geoffrey Khan (Hrsg.)]: Proceedings of the 2nd International Conference of L’Association internationale pour la Dialectologie Arabe held at Trinity Hall in the University of Cambridge 10-14 September 1995. [Cambridge: University Publications Centre – Middle East Centre], S. 259-268. –––– 2006. Das Kairenisch-Arabische. Eine Grammatik. Wiesbaden: Harrassowitz (Porta Linguarum Orientalium, Neue Serie, 22). Yoda, Sumikazu. 2005. The Arabic Dialect of the Jews of Tripoli (Libya). Grammar, Text and Glossary. Wiesbaden: Harrassowitz (Semitica Viva, 35).

The Arabic definite article: A synchronic and historical perspective Barry Heselwood and Janet C. E. Watson, University of Leeds

Abstract Academic research should push boundaries and challenge authorities, however wellknown and however faithfully followed by others. Jan Retsö has never shied away from pursuing controversial ideas, and, through careful scholarship, has successfully challenged many long-held and widely repeated notions, even where, at the time, he may have appeared to be a lone voice. In this paper, we call on phonetic, phonological and cross-linguistic evidence to challenge popular accounts of the Arabic definite article, and show that some of the lone voices of the past may well have been correct in analyses dismissed in later work as unsubstantiated. Jan, we very much hope that you will enjoy this paper. In this paper we argue that the popular account of the phonology of the Arabic definite article in terms of assimilation to a following coronal consonant is not justified. The accepted account holds that the definite article has an underlying phonological form /al/ (or /l/ in some versions) which surfaces as [al] when the following word begins with a non-coronal consonant, but when the following word begins with a coronal consonant the /l/ completely assimilates to the coronal resulting in a geminate coronal consonant: compare for example al-bint [ʔalbint] ‘the girl’ and al-zaffa [ʔaz:af:a] ‘the wedding procession’. We present theoretical and empirical grounds for rejecting the assertion that the /l/ of the definite article assimilates to a following coronal consonant in any synchronically meaningful sense of ‘assimilation’. We argue that for something to count as synchronic assimilation it must be optional, meaning that an unassimilated pronunciation must also be allowed by the grammar. In the absence of counter-evidence, we also assume that optionality implies gradience. Historical assimilation, by contrast, admits neither optionality nor gradience, only unsystematic tokento-token variation in the realisation of the product of a historical process of change. Using illustrative acoustic and electropalatographic data, the situation with the /l/ of the definite article when followed by a coronal consonant is compared to withinword sequences of /l/+coronal consonant in alzam /alzam/ ‘most necessary’, alṭaf ‘most kind’, the form I doubled verbal coronal geminate in hazza /hazza/ ‘to shake’, and the optional assimilation of word-final /l/ to word-initial /r/ in ḥabil rafīʿ ‘a thin rope’. It is also compared with the optional assimilation of the definite article /l/ to a following dorsal stop in Cairo Arabic. Electropalatographic and acoustic data are presented to support the argument that forms such as [az:af:a] should be regarded synchronically not as assimilation but as a

158

Barry Heselwood and Janet C. E. Watson

type of ‘true’ or ‘lexical’ geminate resulting from phonologically-determined allomorphy.

1

The Arabic definite article

The definite article in Arabic forms a syntactic word with the noun or adjective which it defines (Watson 2002: 61-62), and in context forms a phonological word with the preceding syntactic word (ibid). Standard Arabic and the majority of the dialects outside the south-west of the Arabian Peninsula show six allomorphs (Haywood and Nahmad 1965: 22): it has the phonological form /ʔal/ or /ʔil/ when the noun or adjective is utterance-initial and begins with a noncoronal consonant or vowel; it has the form /l/ when the noun or adjective begins with a non-coronal consonant and follows a vowel-final word; when the noun or adjective begins with a coronal consonant, however, the consonant of the article must be that same coronal consonant (/Ccor/), giving /ʔaCcor/ or /ʔiCcor/ in utterance-initial position and /Ccor/ in utterance-medial position following a vowel-final word. In traditional Arabic grammar, the fourteen noncoronal consonants are known as al-ḥurūf al-qamarīya (‘letters of the moon’) (/b ʤ k q ʔ f χ ʁ ħ ʕ h m w j/), and the fourteen coronal consonants as al-ḥurūf alšamsīya (‘letters of the sun’) (/t ṭ d ḍ θ ð ð̣ s ṣ z ʃ l n r/).1 Table 1 provides some examples from Standard Arabic. Gloss

Non-coronal initial consonant

Gloss

Coronal initial consonant

the moon the boy the girl the big book

ʔal-qamar ʔal-walad ʔal-bint

the sun the figs the journey the long river

ʔaʃ-ʃams ʔat-ti:n ʔas-safar

ʔal-kita:bu l-kabi:r

ʔan-nahru ṭ-ṭawi:l

the wedding ʔaz-zaffa procession Table 1: Arabic definite article before a selection of non-coronal and coronal consonants the name

ʔal-ism

Many dialects in the south-west of the Arabian Peninsula, including dialects of southern Saudi Arabia and Yemen, do not exhibit the /l/ definite article in any environment. Several dialects of far northern Yemen and the coastal plain, 1

Emphatic consonants are represented with a subscript dot; /ʤ/ was historically a non-coronal /g/ although it now behaves as a coronal consonant in many modern varieties with respect to the definite article.

The Arabic definite article: A synchronic and historical perspective

159

including Saudi Rijāl Alma‘ (Asiri 2009) and Yemeni Minabbih (Behnstedt 1987: 85) exhibit /m/ ~ /am/ ~ /im/ with no (complete) assimilation to any following consonant, as in: am-safar ‘the journey’, am-qamar ‘the moon’. A selection of dialects in northern Yemen exhibit an /n/ definite article, which again shows no (complete) assimilation to any following consonant, as in northern Yemeni Majz in-ṣaʿbah ‘the female donkey foal’, in-šams ‘the sun’. Dialects scattered throughout the western Yemeni mountain range, exhibit an article which involves gemination of any nominal-initial consonant, as in: ab-bēt ‘the house’, aggamar ‘the moon’, ih-hōd ‘the wedding’ attested, for example, in Rāziḥīt, Jiblah, Ġamar and Xawlān (cf. Behnstedt 1987: 85). Rhodokanakis (1908–1911) noted for some dialects of Oman that nominal-initial labials and uvulars (b, f, q) are geminated when the nominal is defined. Finally, the dialect of the Āl Wahība in the Sharqīya region of Oman shows gemination of nominal-initial coronals, the voiceless velar stop and /j/ (as a reflex of jīm), as in; ak-kabd ‘the liver’, ak-kūʕ ‘the elbow’, ay-yibal ‘the mountain’; l- before laryngeals, pharyngeals and uvular fricatives, as in: al-ḥabal ‘the rope’, al-xabz ‘the bread’, al-ġēl ‘flood water’; luand li- respectively before the glides /w/ and /j/; and a- before labials and /g/, as in: a-bint ‘the girl’, a-midnī ‘the pregnant camel’, a-gabīla ‘the tribe’ (Domenyk Eades p.c.). 2 Review of accounts that take an assimilatory view Most linguistic accounts of the definite article in dialects which exhibit the [ʔVl] ~ [Vl] ~ [l] ~ [ʔVCcor] ~ [VCcor] ~ [Ccor] variants establish an underlying form /al/ (or /l/) with feature-spreading rules, or gestural phasing, to the left of a nominal with an initial coronal (Salib 1981, Watson 2002, Woidich 2006, Youssef 2013). Adopting feature geometry models, Watson (2002) and Youssef (2013) propose that definite article assimilation is motivated by a lexical violation of the Obligatory Contour Principle on the [coronal] tier – both /l/ and the adjacent coronal consonant are marked on the place tier as [coronal]. Assimilation to a following coronal consonant is total – manner and voice, and also place (e.g. ranging from dental /θ ð/ to postalveolar /ʃ/). Youssef (2013: 26) further proposes that assimilation of /l/ of the article to a following coronal results in a false geminate rather than a true geminate. What is interesting about these analyses is the fact that with the exception of /l/ of the definite article, /l/ in Arabic only ever assimilates productively and totally to a following coronal sonorant (Wensinck 1931, Watson 2002, Youssef 2013), rather than to all coronal con-

160

Barry Heselwood and Janet C. E. Watson

sonants, as in Baghdadi /baddal-na/ [baddanna] ‘we changed’, /staʕmal raff/ [staʕmar raff] ‘he made a shelf’ (Youssef 2013).2 3

A non-assimilatory account

In this section, we begin by defining terms we use in this paper: ‘true’ geminate, ‘false’ geminate, ‘fake’ geminate, and assimilation. We will then consider examples of different kinds of phonetic accommodation in Arabic involving /l/ followed by a coronal consonant, and also different kinds of geminates involving coronal consonants, namely: within-word coarticulation, coarticulation across a word boundary, and ‘true’ geminates. Then we will consider which of these phenomena is most like what we observe when the definite article is followed by a nominal (noun or adjective) beginning with a coronal consonant, and by a nominal beginning with a velar stop in Cairene Arabic. Accommodations will be described and analysed in terms of articulatory gestures and gestural phasing (see, for example, Gafos 2002), illustrated with some electropalatographic (EPG) and spectrographic data from two male speakers speaking Modern Standard Arabic. Impressionistic transcriptions of the auditory qualities of the sequences will also be given. Table 2 presents the test words used for this part of the study and the number of tokens collected from each of the two speakers.

2

Sibawayh discusses examples where /l/ may assimilate in Classical Arabic to dentals and interdentals, but states that assimilation of /l/ is less frequent before interdentals than dentals and less frequent before dentals than apical sonorants (Sibawayh II: 416-417, cited in Testen 1998: 151-152).

The Arabic definite article: A synchronic and historical perspective

161

Speakers alsan ‘most eloquent’ alzam ‘most necessary’ alṭaf ‘most kind’ χassa ‘to be mean’ hazza ‘to shake’ ḥaṭṭa ‘to put’ al-sahm ‘the arrow’ al-zaffa ‘the wedding procession’ al-ṭabaq ‘the cover’

A 6 6 6 6 6 6 6 6 6

B 6 6 6 6 6 6 6 6 6

Number of tokens 12 12 12 12 12 12 12 12 12

Table 2. Test words and number of tokens

3.1 Terminology Relevant terms here are geminate types, coarticulation and assimilation. Regarding geminates, we distinguish ‘true’ geminates, ‘false’ geminates, and ‘fake’ geminates. ‘True’ geminates are typically defined as geminates which are ‘monomorphemic and non-derived’ (Davis 2011: 880, fn.7). They are also known in the literature as ‘lexical’ geminates because of their assumed nonderived status in the lexicon (e.g. Oh and Redford, n.d.). However, it is problematic to identify a geminate as ‘true’ or ‘lexical’ in a language exhibiting rootand-pattern morphology such as Arabic in which every occurring major class lexical item is said to be derived from an abstract consonantal root by the application of vowelling patterns, each pattern being the phonological form of at least one morpheme. Our suggestion is therefore to define a ‘true’ geminate as one which is obligatory, contrasts (at least potentially) with singletons, and displays ‘geminate inseparability’ (Gafos 2002: 274) which means that the articulation cannot be released until the end of the geminate; we avoid the term ‘lexical’ geminate. ‘False’ and ‘fake’ geminates emerge on concatenation: ‘false’ geminates through the concatenation of two identical consonants (cf. Oh and Redford, n.d.), e.g. bad dog, and ‘fake’ geminates through assimilation of one consonant to another (usually following) consonant, e.g. bad boy pronounced as [bab˺ bɔɪ]. ‘False’ geminates may also be referred to as concatenative geminates, while ‘fake’ geminates may be referred to as assimilatory geminates. We will use the term ‘coarticulation’ as a generic term for overlapping gestures whether within a word or across a word boundary. The term ‘assimilation’ will be used for cases of coarticulation where the realization shows a com-

162

Barry Heselwood and Janet C. E. Watson

plete change of phonetic category due to the total displacement of one gesture by another, for example when the English phrase in bits is pronounced [ɪm˺ bɪts] with no evidence of a coronal gesture during the pronunciation of in. Non-assimilatory coarticulation is when there is what Jones (1972: 217-21) describes as ‘similitude’, which is due to gestural overlap across the boundary between two segments such that one segment influences the production of the other, or they both influence each other. 3.1 Within-word coarticulation – alzam, alṭaf In the word alzam /ʔalzam/ ‘most necessary’, the /l/ is the first radical of the root which is brought into contact with the second radical as part of the manner of forming elatives on the /ʔafʕal/ pattern (Hayward and Nahmad 1965: 88-90). The first syllable, /ʔal/, is therefore phonologically the same as the form of the definite article when prefixed to a nominal beginning with a non-coronal consonant, e.g. al-bint /ʔalbint/ ‘the girl’. Figure 1a presents EPG frames of the /-lz-/ sequence in alzam produced by a male Libyan speaker from Tripoli speaking Modern Standard Arabic (figure 1b shows the articulatory zones of an EPG frame). These EPG frames, and those in other figures below, are sampled at 10ms intervals. In figure 1 they reveal a gradual reduction in the amount of tongue contact during the /l/ from the frame of maximum contact (frame 218 with 31 contacted electrodes) such that by frame 227 the contact is almost identical, with 24 contacted electrodes, to the contact pattern at the onset of /z/ in frame 228 with 23 contacted electrodes; the difference is that a central channel starts to open up for the fricative in frame 228. During the realization of /z/, perseveration of the lateral pattern of contact can be seen at the back righthand side where there is the same gap for lateral airflow as can be seen during the realization of /l/. The articulation of /z/ here can be symbolized as [ʫ] (IPA 1999: 188), denoting simultaneous lateral and central airflow, although we must assume that lateral airflow is minimal given that the auditory impression is of [z].

The Arabic definite article: A synchronic and historical perspective

163

alveolar postalveolar palatal velar

Figure 1: a) EPG frames of the /-lz-/ sequence in alzam – auditory impression [-lz-]; line indicates segment boundary b) An enlarged blank EPG frame showing the articulatory zones. Right and left correspond to right and left in the speaker’s mouth

In gestural terms, figure 1a shows that the articulatory configurations for realizing /l/ and /z/ overlap to some extent in a real-time dynamic relationship of mutual influence which is strongest at the segment boundary. Figure 2 models coarticulation by representing the phasing of gestures such that gesture B begins (point d) before gesture A has been completed (point f). The interaction of the two gestures is greatest at point e where the amplitude of gesture B equals that of gesture A; we can identify point e in figure 1a between frames 227 and 228. A

normalized

B

amplitude of gesture

c

d

e

f

g

time Figure 2: Gestures phased so as to overlap in time

164

Barry Heselwood and Janet C. E. Watson

In addition to accommodations of the primary articulation gestures involving the tongue tip and lateral margins, there is auditory and acoustic evidence for accommodation of the tongue body as well. The /l/ is realized with a ‘light’ timbre before /z/ in alzam but a ‘dark’ timbre before the emphatic /ṭ/ in alṭaf. The influence of the /ṭ/ can be heard in both vowels which have the low back [ɑ] quality found in realizations of /a/ in the environment of emphatics in Arabic, in contrast to the front [a] quality in other contexts. The emphatic pharyngealization gesture for the realization of /ṭ/ is coarticulated with the gestures for the vowels and for /l/ with acoustic consequences in the form of prominent low-frequency resonances which can be seen in the righthand spectrogram in figure 3. It is well-reported in instrumental studies of Arabic emphatics that the secondary articulation begins earlier and finishes later than the primary articulation (see e.g. Watson 1999).

ʔa

l

z

a

m

ʔ ɑ

lˤ

tˤ

ɑ

f

Figure 3. Spectrograms showing a front [a] vowel and clear [l] in alzam (left) and a back [ɑ] vowel and dark (pharyngealized) [lˤ] in alṭaf (right); dotted ellipsis picks out the thick band of low-frequency resonance responsible for the ‘dark’ timbre

3.2 Coarticulation across a word boundary - ḥabil rafīʿ Arabic displays accommodation of manner of articulation among lingual sonorants such that those lower down the sonority hierarchy optionally assimilate across a word or morpheme boundary to those higher up the hierarchy to form what we label ‘fake’ geminates (see also examples and discussions in Garbell 1958: 326-327, Watson 2002: 237-239, Heselwood, Howard and Ranjous 2011: 6366). The relevant part of the sonority hierarchy is shown in (1). (1)

n

l

r

j

increasing sonority

An example of coarticulation across a word boundary is provided in figure 4 where EPG frames of the /-l#r-/ sequence in three productions of the phrase

The Arabic definite article: A synchronic and historical perspective

165

ḥabil rafīʿ /ħabil rafi:ʕ/ ‘a thin rope’ are shown. The speaker is a female from Syria, speaking the standard Damascus dialect of Syrian Arabic.3 The first production (a) exhibits the same kind of coarticulation as we saw above in alzam. More extensive coarticulation is found in the second one (b), while the third one (c) is an example of complete assimilation resulting in a geminate central approximant [ɹ:]. Each will be described now in some detail. In a), frames 239-243 exhibit full alveolar closure and a gap at the back righthand side for the realization of /l/. In frames 246-252 we can see full lateral closures and a central open channel for an approximant realization of /r/; frames 244-247 show the transition from /l/ to /r/ where the two gestures overlap and influence each other. As in the case of alzam, the two adjacent consonants overlap but maintain perceptual distinctness. In b), the lateral configuration for /l/ rapidly becomes more /r/-like through frames 197-200 while maintaining lateral openings for airflow until frame 204. Gestural overlap can account for this pattern in the same way as we saw above, the difference being that the gesture for /r/ begins earlier resulting in a realization of the sequence in which perception of the lateral is less clear. In terms of the diagram in figure 2, point d is closer to point c in example b) than in example a).

a) Unassimilated /-l#r-/ with a short gestural overlap showing coarticulation4 in frames 244-247 - auditory impression [-l ɹ-]

3 4

Data reproduced from Heselwood et al. (2011: 84). Jones’ (1972) ‘similitude’.

166

Barry Heselwood and Janet C. E. Watson

b) Unassimilated /-l#r-/ but with earlier inception of the gesture for /r/ - auditory impression [- ɹ-] l

c) Totally assimilated /-l#r-/ with no gesture for /l/ - auditory impression [ɹ:] Figure 4. EPG frames showing a) some coarticulation, b) more coarticulation, and c) complete assimilation, in realisations of the sequence /-l#r-/ in ḥabil rafīʿ

The situation is rather different in c) though. In this token, there is no perceptual or articulatory presence of a lateral consonant. The evidence is of complete assimilation with only a [ɹ]-gesture between the final vowel in ḥabil and the first vowel in rafī‘. In terms of the diagram in figure 2, there is no A gesture at all, and thus no point e. The ḥabil rafīʿ examples show a range of degrees of accommodation of the final /l/ to the following initial /r/, from relatively short gestural overlap in a) with a perceptually robust [l], through more extensive overlap in which [l] is less robustly present in articulatory and perceptual terms, to total assimilation in which there is no discernable [l]-segment in production or perception. The result of this total assimilation is a long [ɹ:] extending across the word boundary, our ‘fake’ or ‘assimilatory’ geminate. An important observation in relation to these variable productions is that they demonstrate gradience and optionality. That is to say, the phonology of Arabic allows speakers to use any of these variants at fake geminate sites. A further very important point is that the possibility of releasing the word-final consonant before forming the following word-initial

The Arabic definite article: A synchronic and historical perspective

167

consonant decreases as the degree of gestural overlap increases, reaching zero in the case of total assimilation. A pronunciation such as *[ħabiɹü ɹafi:ʕ] is not attested in Arabic and we have what Gafos (2002: 274) calls ‘geminate inseparability’.5 Similarly in English, if there is assimilation of /n/ in in bits, it cannot be pronounced *[ɪmü bɪts]. A crucial question concerning tokens which exhibit total assimilation is whether the final consonant of the first word has changed its phonological identity: has the /l/ of ḥabil become /r/, or can we justifiably and coherently say that the /l/ is realized as unreleased [ɹ˺] in this context? If /l/ has become /r/, then the realizational possibilities should be the same as for a word-final /r/ followed by a word initial /r/ as in šaǰar rafīʿ /ʃaʤar rafi:ʕ/ ‘a thin tree’: that is to say, if it is claimed that /l/ has become /r/, then it should behave exactly as /r/ behaves. The fact is that it does not. In šaǰar rafīʿ, although there is the option of not releasing the [ɹ] of šaǰar, and thus forming a ‘false’ (concatenative) geminate, there is also the option of releasing it which demonstrates that a ‘false’ geminate does not exhibit geminate inseparability. ‘Fake’ or ‘assimilatory’ geminates do not have this option because the assimilated consonant must be unreleased or remain unassimilated. Following Heselwood, Howard and Ranjous (2011: 95-96), we can represent the difference between ‘false’ and ‘fake’ geminates using Venn diagrams, as in figure 5. Although final /r/ and final /l/ can both be realized as unreleased [ɹ˺], they nevertheless both have unique realizations in their respective sets of possible realizations and are thus disambiguated. This means that while neutralization of the /r/–/l/ opposition may occur phonetically in individual utterances, it does not occur in the phonological system because it is not obligatory. That is to say, in Saussurean terms, not all neutralizations which can be found in parole are neutralizations in langue.6 To ensure that we do not misinterpret them to be so, we must take account of whole sets of possible realizations, not just those we observe on particular occasions.

5 6

We use the tick ü to explicitly denote release of an articulatory constriction, whether audible or not. For Saussure’s langue-parole distinction see Culler (1976: 29-34).

168

Barry Heselwood and Janet C. E. Watson

šaǰar rafīʿ

[ɹ]

ḥabil rafīʿ

[ɹ˺]

[l]

Figure 5. The main realizational possibilities for final /r/ in shaǰar rafī‘ (left circle) and final /l/ in ḥabil rafīʿ (right circle)

3.3 ‘True’ geminates – hazza ‘True’ geminates are geminates that exhibit geminate inseparability and contrast, at least potentially, with singletons (cf. above). Figure 6 presents EPG frames showing the close approximation constriction for the realization of the ‘true’ geminate /zz/ in a token of the Modern Standard Arabic form I doubled verb hazza /hazza/ ‘to shake’ (same speaker as in figure 1).

Figure 6: EPG frames showing the /zz/ coronal articulation in hazza – auditory impression [z:]

The frames in figure 6 show a central channel for airflow narrowing transversely in the alveolar region. The contact pattern remains stable (see frames 38-49), indicating that a single articulatory gesture is executed and maintained between the two vowels. 3.4 Definite article plus coronal consonant – al-zaffa EPG frames for the so-called ‘assimilated’ /l/ of the definite article plus coronal /z/ in al-zaffa [az:af:a] ‘the wedding procession’ can be seen in figure 7. The speaker is the same as in figures 1, 3 and 6.

The Arabic definite article: A synchronic and historical perspective

169

Figure 7. EPG frames for the underlined part of al-zaffa – auditory impression [z:]

The patterns of lingual-palatal contact in figures 6 and 7 are almost identical, and the auditory impressions are of the same long [z:]. This is of course to be expected – nobody suggests that any trace of a lateral articulation can be observed in pronunciations of the definite article when it is followed by a word beginning with a non-lateral coronal consonant such as /z/. The question is therefore not whether there is evidence of [l] in al-zaffa, but whether the observed long [z:] should be analysed phonologically as containing an underlying /l/ which totally assimilates to the following /z/ in manner and aspect of articulation (in the case of words beginning with voiceless /t ṭ θ s ṣ ʃ/, assimilation of glottal state would also have to be postulated, and in the case of /θ ð ð̣ ʃ/ assimilation of place of articulation), or whether it makes more sense to regard the phonological form of al-zaffa as /azzaffa/. Where we find clear examples of assimilation such as in ḥabil rafī‘ pronounced as [ħabiɹ˺ ɹafi:ʕ], we also find the possibility of unassimilated forms exhibiting various degrees of coarticulation consistent with real-time dynamic relationships between adjacent elements also seen between /l/ and /z/ in alzam in figure 1, and between /l/ and /ṭ/ in alṭaf in figure 2. Total assimilation resulting in ‘fake’ geminates across a word boundary, like in ḥabil rafīʿ, is simply the limiting case of the element in the more dominant position exerting its maximum influence in real time but, crucially, without compromising the phonological identity of the assimilated element: /l/ remains /l/ even when realized as [ɹ˺] because of the other members of the set of possible realizations. In words such as hazza with ‘true’ geminates, however, the geminate is not a member of a set of alternative non-geminate realizations, and the same is true of al-zaffa and all other so-called definite article plus coronal consonant assimilations. Our contention is that they are not assimilations at all, but ‘true’ geminates which occur as phonologically determined allomorphs of the definite article.

170

Barry Heselwood and Janet C. E. Watson

Setting up /al/ (or /l/) as the underlying form of the Arabic definite article, and deriving the geminated forms which occur in the context of a following coronal consonant from it through a process of assimilation, fails to account for the lack of optionality and gradience which real-time dynamic accommodations exhibit. There is no evidence at all of a real-time process of accommodation of /l/ to a following coronal of the kind we can observe in ‘false’ gemination when a speaker selects an item starting with a certain consonant and places it after an item ending in that consonant, or a consonant which can assimilate to it as in the example of ḥabil rafī‘ so as to produce a ‘fake’ geminate. Phonetic analysis of constructions such as al-zaffa strongly suggests that speakers do not select /al/ (or /l/) and /zaffa/ and then put them into construction in real time. Rather, it suggests either that they select the syntactic element /azzaffa/ with its geminate /zz/ ‘ready-made’, or that speakers choose the definite article allomorph according to the initial segment of the defined word. 3.5 Definite article plus dorsal stop in Cairo Arabic In modern Cairo Arabic, when a word beginning with /k/ or /ɡ/ is preceded by the definite article, for example il-kalb ‘the dog’, il-gaṛas ‘the bell’, speakers have the option of pronouncing it either with /l/ ([ʔilkalb], [ʔilɡaṛas]) or with a geminated stop ([ʔik:alb], [ʔiɡ:aṛas]) (see Watson 2002: 217-22).7 This optionality could be the same kind as the optionality attending accommodations across a word boundary as in ḥabil rafīʿ, that is to say an optionality in which a range of variants dependent on gestural phasing is implicated, or it could be a straight binary choice between a form with /l/ on the one hand, and a ‘true’ geminate form on the other. This question can in principle be answered by analysis of EPG data to see if there is any evidence of a gesture for /l/. Unfortunately, we do not yet have such data available. 4 Brief review of the historical development of the definite article Part of the motivation for the assimilation analysis of definite article plus coronal consonant sequences is probably acceptance of the view that historically the /l/ assimilated to following coronals and that the underlying form of the article maintains an original /l/ in all contexts, an argument that is supported by Arabic orthography that represents the definite article as (‫ )ﺍاﻝل‬in all contexts. 7

In Cairo Arabic, as in many other dialects, the vowel in the definite article is /i/, not the /a/ of Modern Standard Arabic.

The Arabic definite article: A synchronic and historical perspective

171

Voigt (1998) and Testen (1998) adopt this position, arguing respectively for /al/ deriving from a demonstrative or an asseverative *l particle. There are, however, reasons for doubting this view, and several Semitic philologists have contested the claim that the definite article in Arabic is underlyingly /al/ from a historical point of view. Kuryɫowicz (1972: 131-2) argues that the historically recorded definite article with the principal allomorph /al/ is a relatively recent innovation. This is pursued by Zaborski (2000), who advances the intriguing proposition that the original article had three allomorphs according to number and gender: an nbased article for masculine singular, a t-based article for feminine singular, and an l-based article for plural. Since both /n/ and /t/ regularly assimilate, /l/ was selected as the orthographic form for the Arabic article as the conso-nant least prone to assimilation. On the basis of historical reconstruction and discussion of the definite article in various Semitic languages, Wensinck (1931) and Ullendorff (1965) present cases for regarding the /al/ form of the definite article as resulting from a historical process of dissimilation of non-coronal geminates: the Hebrew definite article is /ha/ with gemination of the nominal-initial consonant except where this is a laryngeal or pharyngeal; before laryngeals and pharyngeals, the form is /hā/ followed by a non-geminate. According to Ullendorff, the basic form of the Hebrew article is lengthening or tenseness with degemination and vowel lengthening before laryngeals and pharyngeals; according to Wensinck, the basic form of the Hebrew article is hā-. In the (vowel-less) Dadanite and Lihyanite inscriptions, the definite article is /h/ before all but laryngeals and pharyngeals, and is assumed (Ullendorff 1965: 635) to have induced gemination of the initial consonant. Before laryngeals and pharyngeals, gemination is dissolved by insertion of either /n/ or /l/, as in: hlḥmq and hlḥmy (Jaussen II, 474, no. 158, cited in Ullendorff 1965: 636; cf. also Macdonald 2000: 40). Noting further that dialects in Oman have been identified as optionally geminating initial labials and uvular consonants (b, f, q) (Rhodokanakis 1908-1911), and that several dialects geminate velar consonants (cf. Watson 2002 and others), Ullendorff argues that the definite article in Hebrew and Arabic may be more closely related than previously thought, that definiteness was originally expressed by gemination or

172

Barry Heselwood and Janet C. E. Watson

tensing of the initial consonant, with degemination of glides, laryngeals and pharyngeals through /l/.8 Ullendorff’s argument for an original geminate definite article was later dismissed by Wagner (1993) on the basis of evidence from Modern South Arabian. Wagner argued that Mehri showed no morphological gemination, and therefore that gemination could not have been an exponent of definiteness. There are, however, no grounds for Wagner’s dismissal: work has since shown that both Omani Mehri and its sister languages Śḥerɛ̄t and Ḥarsūsī do exhibit gemination of the initial C in definite nouns and adjectives, particularly where this C is voiceless and non-emphatic, as in Mehri: tōmar ‘dates’ > (a)t-tōmar ‘the dates’, xīl ‘uncle’ > (a)x-xaylī ‘my uncle’ (Watson 2012: 20-22). 5 Conclusion We have argued that the geminates which occur in definite article plus coronal consonant constructions are not the result of synchronic assimilation and should instead be regarded as ‘true’ geminates, not assimilatory geminates. Our illustrative articulatory and acoustic data indicate that the geminate [z:] in alzaffa is no different from that in hazza, and very different from within-word coarticulations involving /l/ and from ‘fake’ geminates resulting from assimilation of /l/ across a word boundary. Whether the Arabic definite article began historically with a ubiquitous /l/ which then assimilated to coronals, or whether /l/ was part of an alternation {/n/~/t/~/l/} based on gender and number, or whether the marker of definiteness began as gemination and then developed an /l/-form in a process of historical dissimilation, makes no difference to our analysis, although of course it does make a difference to an account of how the current state of affairs arose. The non-assimilatory account of the phonology of the Arabic definite article we have presented is better accommodated in a theoretical approach which does not assume invariant underlying phonological forms of morphemes from which observable variants are derived. The assumption found i.a. in generative phonology that a single morpheme must at some ultimate abstract level be instantiated by a single phonological form can be characterized as an example of a reification fallacy in which a single form in morphology is required to cor8

We have seen this above for the dialect of Āl Wahība in the Sharqiyya region, where the l- allomorph is restricted to glides, and laryngeal, pharyngeal and uvular fricatives.

The Arabic definite article: A synchronic and historical perspective

173

respond to a single item in phonology. Setting up an underlying form with /l/ is very likely motivated by one or more of three factors. First is the fact that in written Arabic it is represented with the letter corresponding to /l/ in all orthographic contexts, as mentioned above. Second is the widely-held view that the original form of the article had /l/ in all phonological contexts. Third is the fact that, in formulating a phonological rule to derive surface forms, it is relatively simpler to derive the geminate variants from an underlying /l/ because they can all be specified by the single feature-value [+coronal] with the [l] variant occurring ‘elsewhere’ (see Kiparsky 1973). Taking each of these motivations in turn: regarding the first point, it is often risky to base phonological analysis on orthographic evidence; regarding the second point, the original form of the definite article may or may not have had an /l/, the arguments for and against remain inconclusive; as for the third point, there is little if any justification for assuming that the simplest rule we can come up with bears any relation to what language users actually do in real time when speaking. To our knowledge, there is no evidence from available data that Arabic speakers, when saying al-zaffa, begin with /ʔal-zaffa/ and then assimilate the /l/ to the /z/. If that was the case, we would expect to find real-time dynamic influences of the kind seen in alzam, alṭaf and ḥabil rafī‘ as the articulatory gestures for realizing /l/ adapt to the local circumstances. In our opinion, the facts of the definite article in Arabic are best accounted for in terms of phonologically conditioned allomorphy, not by derivation from a single invariant form. There is an interesting further consequence of our analysis for the notion of a ‘true’ geminate which we only have space to touch on here, but which we would nevertheless like to point out. It relates to the point made above about the problematic nature of the definition of a ‘true’ geminate as monomorphemic and non-derived. In a case such as al-zaffa the definite article morpheme has been prefixed to a noun, resulting in an obligatory, non-gradient and inseparable geminate [z:]. It is clearly not monomorphemic, and is clearly derived in the sense of coming about due to prefixation. Similarly, in a case such as kassara ‘to smash (s.th.)’ in which the obligatory non-gradient inseparable geminate [s:] is classed as a ‘true’ geminate, a morpheme with the function of intensifying a verb has been affixed to kasara ‘to break (s.th.)’. kassara is thus grammatically derived from kasara by addition of a morpheme. There are therefore strong grounds for regarding these two geminates as belonging to the same type in Arabic. Our contention is that they are both ‘true’ geminates in the revised

174

Barry Heselwood and Janet C. E. Watson

typology of geminates presented in Table 3, which does not appeal to derivational relations as a criterion. TRUE

hazza kassara al-zaffa Inseparable Obligatory Potential contrast with singleton

FALSE

FAKE

šaǰar rafīʿ

ḥabil rafīʿ

Eng. bad dog

Eng. bad boy

PSEUDO

Coll. hazz ü ü

x x

ü x

ü ü

ü

ü

ü

x

Table 3. A revised typology of geminates

If kassara contains a true geminate, then so do al-zaffa, al-šams, al-tīn, al-nahr and all words containing the definite article followed by a coronal consonant. The category of ‘pseudo’ geminate covers those final long consonants in colloquial forms of Arabic resulting from vowel apocope, giving e.g. /hazz/ compared to MSA /hazza/. If there is no possibility of commutation with a singleton */haz/ because of syllable weight requirements for monosyllabic words, then the long consonant represents a neutralization between geminate and singleton for which ‘pseudo geminate’ may be an appropriate and useful term. However, further research is needed to confirm whether such neutralization does in fact occur in Arabic. The notion of pseudo-geminate does seem to be relevant in the kind of quantity complementarity found in Swedish where monosyllabic words either have a short vowel and long coda consonant, or a long vowel and a short coda consonant, e.g. hatt [hat:] ‘hat’, vs. hat [ha:t] ‘hate’ (Witting 1977). There is no possibility of a short-long consonant contrast after a short or long vowel, and no possibility of a short-long vowel contrast before a short or a long coda consonant. We propose that the long [t:] in [hat:] is an example of a pseudo-geminate consonant.

The Arabic definite article: A synchronic and historical perspective

175

References Behnstedt, Peter. 1987. Die Dialekte der Gegend von Ṣaʿdah (Nord-Jemen). Wiesbaden: Harrassowitz. Culler, Jonathan. 1976. Saussure. Fontana Paperbacks. Davis, Stuart. 2011. “Geminates.” In Marc van Oostendorp, Colin J. Ewen, Elizabeth Hume and Keren Rice (eds). The Blackwell Companion to Phonology Volume II: Suprasegmental and prosodic phonology, 873-897. Malden MA: Wiley-Blackwell. Gafos, Adamantios. 2002. “A grammar of gestural coordination.” Natural Language and Linguistic Theory 20: 269-237. Garbell, Irene. 1958. “Remarks on the historical phonology of an East Mediterranean Arabic dialect.” Word 14: 303-337. Haywood, John A. and H. M. Nahmad. 1965. A New Arabic Grammar. London: Lund Humphries. Heselwood, Barry, Sara Howard and Rawya Ranjous. 2011. “Assimilation of /l/ to /r/ in Syrian Arabic.” In Zeki Majeed Hassan and Barry Heselwood (eds). Instrumental Studies in Arabic Phonetics, 63-98. Amsterdam: John Benjamins. IPA. 1999. Handbook of the International Phonetic Association. Cambridge: Cambridge University Press. Jones, Daniel. 1972. An Outline of English Phonetics. Cambridge: Cambridge University Press. 9 edition. Kiparsky, Paul. 1973. “‘Elsewhere’ in phonology.” In Stephen Anderson and Paul Kiparsky (eds). A Festschrift for Morris Halle, 93-106. New York: Holt, Rinehart & Winston. Kuryɫowicz, Jerzy. 1972. Studies in Semitic Grammar and Metrics. Wroclaw: Ossolinuem. Macdonald, Michael. 2000. “Reflections on the linguistic map of pre-Islamic Arabia.” Arabian Archaeology and Epigraphy 11: 28-79. th

Oh, Grace E. and Melissa A. Redford (n.d.) “The production and representation of fake geminates in English.” http://pages.uoregon.edu/redford/Papers4download/ Oh&Redford_v2.pdf (accessed 28 November 2013). th

Rhodokanakis, Nikolaus. 1908-1911. Der Vulgärarabische Dialekt im Dofār (Zfār). Wien: Österreichische Akademie der Wissenschaften. Sibawayh, Abū Bišr ʿAmr b. ʿUṯmān b. Qanbar. 1965/1317. Kitāb Sībawayhi. Beirut: alMaṭbaʿah al-Kubrā. Testen, David. 1998. Parallels in Semitic Languages. Leiden: Brill. Ullendorff, Edward. 1965. “The form of the definite article in Arabic and other Semitic languages.” In George Makdisi (ed.). Arabic and Islamic Studies in Honor of H. A. R. Gibb, 631-637. Leiden: Brill. Voigt, Rainer. 1998. “Der Artikel im Semitischen.“ Journal of Semitic Studies 43: 221-258. Wagner, Ewald. 1993. „Gedanken zum Verb des Mehri aufgrund der neuen Materialien von Johnstone.“ Zeitschrift für Arabische Linguistik 25: 316-339. Watson, Janet C. E. 1999. “The directionality of emphasis spread in Arabic.” Linguistic Inquiry 30: 289-300.

176

Barry Heselwood and Janet C. E. Watson

Watson, Janet C. E. 2002. The Phonology and Morphology of Arabic. Oxford: Oxford University Press. Watson, Janet C. E. 2012. The Structure of Mehri. Wiesbaden: Harrassowitz. Wensinck, Arent J. 1931. “The article of determination in Arabic.” Mededeelingen der Koninklijke Akademie van Wetenschappen Deel 71, Serie A. Witting, Claes. 1977. Studies in Swedish Generative Phonology. Stockholm: Almqvist & Wiksell. Youssef, Islam. 2013. Place assimilation in Arabic: Contrasts, features, and constraints. Ph.D Thesis, University of Tromsø. Zaborski, Andrzej. 2000. “Inflected article in Proto-Arabic and some other West Semitic languages.” Asian and African Studies 9: 24-35.

The position of Mardin Arabic in the Mesopotamian-Levantine dialect continuum Otto Jastrow, Tallinn University

I.

Introduction

In May 2013 I participated in a conference convened by Mardin University under the title 1st International Symposium on Spoken Arabic Dialects and their Oral Literature in Turkey. Mardin plays an important role in my own biography since it was the place of my first fieldwork, two periods of three months in 1967 and 1968 – almost a lifetime ago, but apart from these private considerations I am convinced that Mardin Arabic is crucial for the understanding not only of the Anatolian Arabic dialect landscape but also of the development of the sedentary dialects of both Syria and Iraq. I thus found it very suitable to speak about the importance of Mardin Arabic in the very place where it is spoken. My presentation was called Arabic Dialects of South-Eastern Anatolia and the Role of Mardin Arabic, and I even ended it with a short passage in Mardin Arabic. Nevertheless the audience showed almost no reaction, which left me somewhat disappointed. In retrospect, however, the reason for this lack of response became clear to me. The majority of the audience was made up of professors from all over Turkey, most of them Turkologists who did not know any Arabic and, more often than not, only little English. Arabic dialectology as an academic discipline does not exist in Turkey and therefore was probably alien to most of them. So my paper simply did not reach its audience. This prompted me to present the same paper in a shortened version under the title What is so special about Mardin Arabic? at the AIDA Symposium on Arabic dialectology at the University of Bergen, October 10-11, 2013. This time, of course, the paper met with decidedly greater interest and provoked some discussions. Some of the comments voiced on that occasion I have gratefully incorporated in the present revised and enlarged version. It is not without some embarrassment that I dedicate this paper which is of a preliminary character to my distinguished colleague Jan Retsö who in his work has demonstrated how one can think an idea to its very end.

178

Otto Jastrow

II. The Anatolian qәltu dialects On the territory of the Turkish Republic a considerable number of Arabic dialects are spoken to this day. The following paragraphs concentrate on one particular dialect group, the so-called Anatolian qәltu dialects which, unlike all other Arabic dialects spoken in Turkey, are primarily affiliated with Mesopotamian Arabic but, as we shall try to show in the course of our discussion, at the same time mark a point of transition in the Mesopotamian-Levantine dialect continuum. Qәltu dialects are still spoken in a number of places in the vilayets of Mardin and Siirt, including the capitals of these two vilayets, and in the Kozluk-Sason area. Almost all of the present-day dialects are spoken exclusively by Muslims because the Christian speakers were mostly killed during the Armenian holocaust in 1915, and the remaining Christians have been pushed out of the country ever since. The Jews, in turn, left Turkey to settle in Palestine or, after its foundation, in the State of Israel. It is a fair estimate that due to those historic events about half of the qәltu dialects of old have died out although data for some of them could be salvaged in the last minute. The following table (1) shows the distribution of qәltu dialects in Turkey – not as they appear today but as they existed 100 years ago; dialects given in italics are now extinct or have disappeared due to forced emigration. (1) Anatolian qǝltu dialects I. Mardin group Mardin town (Muslims; Christians mostly emigrated) Mardin villages (Muslims; Christians emigrated) [Bәnēbīl (chr.), Qalʿәt maṛa, әṣṢawr] Plain of Mardin (Muslims; Christians extinct) [Ḥәrrīn, Salāx, Tall Šʿēr, Tall Fayḏ̣ – әlManṣūrīye (Christians, extinct)] Kōsa and Mḥallami dialects (Muslims, 1 Christian village) [Kōsa, ca. 15 villages, e.g. Arbәl, Ǧawze] [Mḥallami, ca. 20 villages, e.g., Kәndērīb, 1 Christian: Qәllәf] Āzǝx (Christians, now emigrated) II. Siirt group Siirt town (Muslims; Christians extinct) Siirt villages (Muslims) [6 villages: Tәllo, Fәrsāf, Ḥalanze, Snēb, Tōm, Fәsken]

The position of Mardin Arabic

179

III. Diyarbakır group Diyarbakır town (Christians, extinct; Jews, emigrated) Diyarbakır villages (Christians, extinct) Diyarbakır, Siverek, Čermik, Urfa (Jews, emigrated) IV. Kozluk-Sason-Muş group Kozluk (Muslims; Christians extinct?) [Daragözü (Muslims)] Sason (Muslims; Christians extinct?) [Hasköy (Muslims)] As can be seen from the above table, the qәltu dialects formerly spoken by Christians and Jews, with very few exceptions, have disappeared from the map and almost all of the surviving dialects today are spoken by Muslims. To the panorama of these dialects which have been known for some time, in particular due to the works of the present author1, a new variety was added as recently as in 2008 when Shabo Talay discovered a new dialect spoken in the village of Sine   (Turkish: Oyuklu) in the district of Lice, Diyarbakir province.2 The importance of this new dicovery for the whole field has still to be fully elaborated, however, for the present article it leads to some important conclusions. III. Mardin Arabic3 as the most conservative Anatolian qәltu dialect The present paper argues that of all qәltu dialects, both surviving and extinct, Mardin Arabic or Mērdәlli, the dialect spoken in the town of Mardin, is the most important one from a dialectological point of view. As will be shown, it is the most conservative dialect of the whole qәltu group and therefore can serve as a starting point to explain the developments which have taken place in the other dialects and led to their characteristic features. One of the most striking examples for the conservativism of Mardin which at the same time illustrates the differences among the qәltu dialects is the deve1

Cf. Jastrow 1978, 1981, 2003, 2005.

2

Cf. Talay 2012

3

This paragraph deals with the differences between the dialect of Mardin proper and the various subdialects of the Mardin group. Thus „Mardin“ here is used strictly for the dialect of the town whereas in the following paragraphs it can stand both for the dialect of the town and the subdialects, in other words, the whole Mardin group.

180

Otto Jastrow

lopment of the Old Arabic interdentals, as shown in table (2). As elsewhere in the Arab world, the Old Arabic consonants ḍād and ẓā have been merged into a single consonant which was pronounced as a voiced interdental emphatic, *ḏ̣. Therefore there are three interdentals whose development we have to trace, voiceless *ṯ, voiced *ḏ and, as pointed out, voiced emphatic *ḏ̣. The following table summarizes the sound shifts taking place in the main Anatolian qәltu subgroups. (2) Different reflexes of the interdentals Mardin Āzǝx Siirt Daragözü *ṯ ṯāṯe sāse fāfe sāse *ḏ axaḏ axaz axav aġaz ẓarab ṿarab ẓarab *ḏ̣ ḏ̣arab

Diyarbakır tlāte axad ḍarab

“three” “he took” “he shot”

What makes this list so interesting is the fact that on the relatively small territory of the qәltu dialects there occur all the possible reflexes of the interdentals found in the whole Arab world. Within the qәltu group it is only Mardin (and some dialects of the Mardin subgroup, such as Kōsa and Mḥallami) which have preserved the original pronunciation of the interdentals. Another phonetic feature which shows Mardin to be more conservative than most of the surrounding qәltu dialects concerns the pronunciation of the voiced pharyngal fricative, ʕayn. In the Mardin dialect ʕayn has been preserved throughout, irrespective of its position in the word. In most of the remaining qәltu dialects, however, ʕ has become voiceless, in other words ḥ, when it occurs either word finally or word internally before a voiceless consonant, as shown in table (3): (3) Devoicing of *ʕ Mardin Āzǝx bāʕ, ybīʕ bāʕ, ybīʕ bǝʕtu bǝʕtu

Siirt bāḥ, ybīḥ bǝḥtu

Daragözü Diyarbakır bāḥ, ībīḥ bāḥ, ybīḥ bǝḥtu bǝḥtu

“to sell” “I sold”

The voiced pronunciation is restored, or rather retained, if the original *ʕ is followed by a vowel, thus, e.g., Siirt bāḥ - bāʕu “he sold – they sold”, ybīḥ - ybīʕū “he sells – they sell”.

The position of Mardin Arabic

181

Furthermore in Siirt, word final ḥ, harking back to both *ḥ or *ʕ, can be reduced to zero, especially in high frequency words, e.g., ṛā, yṛō “to go”, ṭallo, yṭallo “to look” – compare Mardin ṛāḥ - yrōḥ, ṭallaʕ - yṭallәʕ. Finally, a third phonetic feature which shows Mardin to be more conservative than the surrounding dialects is the preservation of the initial consonant h of some very common pronouns and adverbs, as shown in table (4): (4) Treatment of initial *h in pronouns and adverbs Mardin hāḏa hawn(e)

Āzǝx hāza hawn(e)

Siirt āva awne

Daragözü Diyarbakır āza āda ōn(i) awne

“this” “here”

To sum up: We have identified three phonetic features which show the dialect of Mardin as more conservative than the surrounding qәltu dialects. Moving on to morphology, it is well known that the qәltu dialects as a whole display a remarkable conservatism, especially in verb morphology, side by side with some innovations. Here again, it will be shown that Mardin retains the most conservative paradigm while the remaining qәltu dialect groups have reshuffled the stem formation of the verb. The conservative character of verb inflection in all the Anatolian qәltu dialects is illustrated by the paradigm of Mardin – table (5): (5) Perfect and imperfect inflection of stem I (Mardin) Perfect

Imperfect

Perfect

Imperfect

3 SG M

katab

yәktәb

šәṛәb

yәšṛab

3 SG F

katabәt katabu

tәktәb yәktәbūn

šәṛbәt šәṛbu

tәšṛab yәšṛabūn

katabt katabti katabtәn

tәktәb tәktәbīn tәktәbūn

šәṛәbt šәṛәbti šәṛәbtәn

tәšṛab tәšṛabīn tәšṛabūn

katabtu

aktәb

šәṛәbtu

ašṛab

katabna

nәktәb

šәṛәbna

nәšṛab

3 PL C 2 SG M 2 SG F 2 PL C 1 SG 1 PL

The paradigm displays the merger of masculine and feminine plurals to a common form (marked as C. = communis generis). This is a wide-spread feature of

182

Otto Jastrow

sedentary dialects; otherwise, however, the inflectional paradigm is fairly conservative. Most striking is, of course, the inflectional ending -tu for the 1. person sg. m. – katabtu, šәṛәbtu – which has given the whole dialect group its name: qәltu dialects. Another remarkable feature is the preservation of the so-called “long imperfect forms“ – yәktәbūn, tәktәbīn, tәktәbūn – , typical for Mesopotamian Arabic. If the inflection of the strong verb is already fairly conservative, the picture becomes even more striking when we look at the inflection of the verbs with final weak consonants (verba tertiae infirmae), see table (6): (6) Some inflected forms of the verba tertiae infirmae Classical Arabic Imperfect 3 PL M/C strong verbs yaktubūna verba tert. inf. yansawna 2 SG F strong verbs taktubīna Perfect

3 PL M/C

verba tert. inf. strong verbs verba tert. inf.

tansayna katabū banaw

Mardin yәktәbūn yәnsawn tәktәbīn tәnsayn katabu banaw

The verbal forms of Mardin Arabic thus correspond largely to the verbal paradigm of Classical Arabic, except for one innovation: the imperfect forms which end in -ī have been restructured in analogy to the ones ending in -ā, as shown in table (7): (7) Analogical restructuring of imperfect verbs ending in -ī *yabnī, *yabnūna *tabnī, *tabnīna

> >

yәbni, yәbnawn (cf. yәnsawn) tәbni, tәbnayn (cf. tәnsayn)

These Mardin forms are shared by the other Anatolian dialects which have, however, introduced some innovations in the formation of the derived verb stems. In the stems II, III, V, VI and X the last syllable of the perfect base does not show the vowel a as in Classical Arabic – and Mardin Arabic – but a raised vowel ә, e or o, see table (8). The scarcity of the data does no allow the use of the same verbs in each dialect to illustrate this phenomenon.

The position of Mardin Arabic

183

(8) Formation of stems II, III, V, VI, X Mardin II ḥammal

Āzǝx ḥammal

Siirt ʕallem

Daragözü Diyarbakır xallǝṭ xalleṭ

III dēxal

dēxal

ʕayyoṭ dēxel

dēxǝl

dēxel

“to scream” “to insert”

X staxbar stanḏ̣ar

staxbar

*staxber

staxbǝr

staxber

“to ask”

stanṿor

“to wait”

How can we explain this development? In the stems concerned the vowel of the final syllable of the imperfect base has been extended by analogy to the perfect, thus, e.g., *ʕallam, yʕallәm (as preserved in Mardin) > ʕallәm, yʕallәm. The short vowel ә in the final syllable is always elided before a vowel suffix, e.g., Siirt ʕallәm + -ū > ʕallmu “they taught”, as against Mardin ʕallamu. In simply closed final syllables, the vowel ә has been preserved with its original quality [ә] in Daragözü while in Diyarbakir it has been raised to [e]. In Siirt, on the other hand, the vowel appears either as [e] or as [o], depending on the phonetic environment, thus, e.g., dēxel “he inserted, entered”, stanṿor “he waited”. A final observation concerns verb syntax. Mardin Arabic uses the simple imperfect to express both a general and a progressive present, e.g., yәktәb “he writes/he is writing” while most other dialects, both inside the Mardin subgroup and in the other subgroups, use a prefix kū- to express the progressive present, e.g., Kәndērīb yәktәb “he writes” vs. kūyәktәb “he is writing”.4 Thus here again, Mardin town emerges as the more conservative dialect. This short summary would be incomplete without mentioning an important innovation which is shared by all qәltu dialects of Anatolia but is not found in the qәltu dialects of Iraq. The feature in question is the so-called copula, as shown in table (9). Note that Siirt shows a different paradigm which puts the copula in front of the predicate.

4

For the etymology of kū- cf. Jastrow 2013.

184

Otto Jastrow

(9) Kopula (“he is, she is, they are“) 3 SG M

Mardin fә-lbayt-we

Āzәx fә-lbayt-u

Siirt ūwe fә-lbayt

”he is in the house”

3 SG F

hawne-we fә-lbayt-ye

hawne-we fә-lbayt-i

ūwe awne īye fә-lbayt

”he is here” ”she is in the house”

hawne-ye

hawne-ye

īye awne

”she is here”

fә-lbayt-әnne hawne-nne

fә-lbayt-әn hawne-nәn

әnne fә-lbayt әnne awne

”they are in the house” ”they are here”

3 PL C

To sum up: Mardin is a mainstream dialect because it has not undergone any extravagant innovations; as a mainstream dialect it is, however, extremely conservative. Compared to Mardin, the remaining qәltu dialect subgroups conform to the concept of typical “peripheral dialects”: They are separated from the continuum of the Arabic language area, therefore they show a strong influence of other, Non-Arabic, regional languages and, having been abandoned on their own, they have gone through some very progressive internal developments. The dialect of Mardin as the most conservative dialect type, can thus be described as the core of the qәltu dialects, surrounded by more progressive dialects: Āzǝx, Siirt, Kozluk-Sason-Muş, Diyarbakır. IV. Anatolian Arabic as a branch of Mesopotamian qәltu Arabic The town of Mossul in Northern Iraq has a dialect very similar to the one of Mardin. This applies to all three varieties, spoken by Jews, Christians and Sunni Muslims, which are distinguished only marginally (Jastrow 2004b). When I first met with people from Mossul, both Muslims or Jews, I could converse with them easily, based on the fluent Mardin Arabic which I had acquired during six months of residence. Albert Socin, the first European scholar to explore the linguistic landscape of South-East Anatolia and Northern Iraq, even called his paper from the year 1882/1883, published in Zeitschrift der Deutschen Morgenländischen Gesellschaft, ”Der arabische Dialekt von Mosul und Märdin“. His view that the two dialects are identical is, however, untenable from a dialectological point of view. Strangely enough, Socin did not mention the most prominent phonetic difference which strikes the listener immediately, namely the shift of *r/ṛ to a postvelar fricative /ġ/ in Mossul, e.g. (Jastrow 1979: 38):

The position of Mardin Arabic

185

(10) shift *r/ṛ > /ġ/ Mardin ṛāḥ

Mossul ġāḥ ”he went“

maṛa kәbәr

maġa kәbәġ

”woman“ ”he grew“

Socin simply used for transcribing both dialects. Another prominent sound change is the monophthongization of *aw and *ay, preserved in Mardin, to the monophongs /ō/ and /ē/, e.g., (11) monophthongization of *aw and *ay Mardin bayt zawǧ

Mossul bēt ”house“ zōǧ ”husband“

The most important structural difference, however, which separates Iraqi and Anatolian qәltu dialects, is of a morphological nature. It concerns the final consonants /-n/ vs. /-m/ in five pronouns and inflectional suffixes as illustrated in table (12): (12) Different final consonants in Anatolian und Iraqi Arabic morphemes Mardin ǧәbtәn әntәn

Mossul ǧәbtәm әntәm

”you (2 PL) brought” ”you (2 PL)”

baytkәn hәnne baytәn

bētkәm hәmme bētәm

”your (2 PL) house” ”they” ”their house”

It is the difference between final -n and final -m in these morphemes which follows exactly the Turkish-Iraqi border and allows us immediately to determine whether a specific dialect is an Iraqi or an Anatolian dialect. Apart from these points where Iraqi and Anatolian qәltu dialects diverge – and which we stressed in order to determine the position of Mardin Arabic – the amount of common features, both in phonology and morphology is, of course, much more important. Without going into details, let us just resume a few shared features:

186

–

Otto Jastrow

Phonology: preservation of /q/, /ǧ/ (as an affricate), the interdentals (with

/ḏ̣/ being pronounced as a voiced velarized interdental fricative), the introduction of new consonant phonemes like /p/, /č/ and /ž/, merging of *i and *u to /ә/, the ʾimāla of *ā > ē, triggered by an *i,* ī in the adjacent syllable. –

Morphology: preservation of the inflectional ending -tu for the 1.sg. perfect

(qәltu, hence the name of the whole dialect group), preservation of the ”long“ inflectional endings of the imperfect (tnāmīn ”you (SG. F.) sleep“, tnāmūn ”you (PL. C.) sleep“, ynāmūn ”they sleep“), preservation of different inflectional endings for strong verbs and verba tertiae infirmae (see III above), e.g., tәnsayn/ tәnsēn ”you (SG. F.) forget, yәnsawn/yәnsōn ”they forget“, conservative forms of the pronominal suffixes (abūhu ”his father“, abūki ”your (SG. F.) father“). Thus if we take into account all prominent linguistic features there can be no doubt that in a broader dialect classification Mardin Arabic as well as the remaining dialects of its group, form a special branch not of Iraqi, but of Mesopotamian qәltu Arabic. V. Mardin Arabic as a pivotal point in the Mesopotamian-Levantine dialect continuum The town of Mardin is not separated from the Arabic language area but is geographically very close to neighbouring Syria. This is seemingly in contrast to the fact that, dialectologically speaking, its dialect – as well as the surrounding varieties mentioned in III above – are part of the Mesopotamian, not the Greater Syrian dialect area. In a number of points, however, the Mardin dialect group represents an area of transition between these two major dialect areas. This thesis is discussed in detail in a forthcoming article by Shabo Talay (Talay, to appear). The author shows that a number of typical qәltu features – e.g., the preservation of /q/, the partial preservation of the diphthongs /aw/ and /ay/, the iconditioned ʾimāla – extend into the Çukurova and Hatay areas of Syrian Arabic (both situated on Turkish territory), whereas the ending /-n/ for a number of morphemes in Mardin Arabic, as discussed in IV above, can be viewed as an extension of a typical Syrian feature, e.g.,

The position of Mardin Arabic

187

(13) Final consonants -m/-n in Syrian, Anatolian and Iraqi Arabic morphemes Hatay (ǧibtu)

Mardin ǧәbtәn

Mossul ǧәbtәm

”you (2 PL) brought”

(intu)

әntәn baytkәn

әntәm bētkәm

”you (2 PL)” ”your (2 PL) house”

hәnne baytәn

hәmme bētәm

”they” ”their house”

baytkin/kun5 hinni baytin/un

The recently discovered dialect of Sine in the district of Lice, province of Diyarbakir (Talay 2012) is one more missing link in this dialect continuum. In phonology it preserves the original /q/ and the diphthongs /aw/ and /ay/ but has shifted the interdentals to sibilants. In morphology, all while preserving the most characteristic morphological feature of the Mardin type, e.g., the inflectional ending -tu for the 1.SG. perfect (qraytu “I read“), and even the copula, Sine shows, on the other hand, short imperfect forms of the type tqūli, tqūlu, yqūlu. Even more important, Sine shows some diachronic developments in the vowel system which are alien to the Anatolian qәltu type but typical for Syrian Arabic. Thus, e.g., long /ā/ in closed final syllable is shifted to /ō/, in the vicinity of back or emphatic consonants, e.g., nōṛ ”fire“, nhāṛōt ”days“, but kān ”he was“, ġalmāt ”goats“. Word final -a, according to Talay (to appear) has undergone two subsequent shifts. In a first step it has been split into /-e/ after front consonants and /-a/ after back or emphatic consonants. In a second step final /-a/ has been shifted to /-o/. Talay adduces examples from the numbers of 110, e.g., *xamsa > xamse ”five“, *sitta > sәtte ”six“ but: *aṛbʿa > *aṛbʿa (cf. Mardin aṛbʿa) > aṛbʿo ”four“ *ʿašṛa >*ʿašṛa (cf. Mardin ʿašṛa) >ʿašṛo ”ten“ Another typical development which Talay discusses in detail is the shift of /a/ in simply closed final syllables to /o/ in back or emphatic consonant environments, e.g., *baydar > *baydaṛ (cf. Mardin baydaṛ) > baydoṛ „threshing floor“. The dialect of Sine whose detailed description by Shabo Talay is still in work represents thus another important step in the Mesopotamian-Levantine

5

The morphemes -kin, -in are characteristic for the Alevi dialects, the morphemes -kun, -un for the Sunni dialects (Arnold 1998: 101).

188

Otto Jastrow

continuum for which the pivotal role of Mardin Arabic has been discussed in this paper. Unfortunately in the area of Kozluk, Hasköy, Mutki and Lice almost all Arabic dialect have been wiped out. Otherwise the gradual shifting of the Mesopotamian dialect type to the Levantine one could be traced in even greater detail.

Literature Arnold, Werner. 1998. Die arabischen Dialekte Antiochiens. Wiesbaden: Harrassowitz. Behnstedt, Peter. 1997. Sprachatlas von Syrien. Kartenband (= Semitica Viva 17). Wiesbaden: Harrassowitz. Grigore, George. 2007. L’arabe parlé à Mardin – monographie d’un parler arabe périphérique. Bucharest: Editura Universităţii din Bucureşti. Jastrow, Otto. 1973. Daragözü. Eine arabische Mundart der Kozluk-Sason-Gruppe (Südostanatolien). Grammatik und Texte. Nürnberg: Hans Carl. Jastrow, Otto. 1978. Die mesopotamisch-arabischen qǝltu-Dialekte. Bd. I: Phonologie und Morphologie (= Abhandlungen für die Kunde des Morgenlandes 43,4). Wiesbaden: Steiner. Jastrow, Otto. 1979. ”Zur arabischen Mundart von Mossul.“ In: Zeitschrift für arabische Linguistik 2: 36-75. Jastrow, Otto. 1981. Die mesopotamisch-arabischen qǝltu-Dialekte. Bd. II: Volkskundliche Texte in elf Dialekten (= Abhandlungen für die Kunde des Morgenlandes 46,1). Wiesbaden: Steiner. Jastrow, Otto. 2001. ”Aramäische Lehnwörter in den arabischen Dialekten der SüdostTürkei.“ In: Stefan Wild and Hartmut Schild (eds.). Akten des 27. Deutschen Orientalisten-tages, 615-621. Würzburg: Ergon. Jastrow, Otto. 2003. Arabische Texte aus Kinderib (= Semitica Viva 30). Wiesbaden: Harrassowitz. Jastrow, Otto. 2004a (appeared 2006). ”Aramäische Lehnwörter im arabischen Dialekt von Kinderib (Südost-Türkei).“ In: Estudios de dialectología norteafricana y andalusí 8. Homenaje a Peter Behnstedt, 99-103. Zaragoza: Instituto de Estudios Islámicos y del Oriente Próximo. Jastrow, Otto. 2004b. “Jüdisches, christliches und muslimisches Arabisch in Mossul.“ In: Martine Haak, Rudolf de Jong and Kees Versteegh (eds.). Approaches to Arabic Dialects. A Collection of Articles presented to Manfred Woidich on the Occasion of his Sixtieth Birthday, 135-150. Leiden/Boston: Brill. Jastrow, Otto. 2005. Glossar zu Kinderib (Anatolisches Arabisch) (= Semitica Viva 36). Wiesbaden: Harrassowitz. Jastrow, Otto. 2006a. “Anatolian Arabic.“ In: Kees Versteegh (ed.). Encyclopedia of Arabic Language and Linguistics. Vol. I, 86-96. Leiden/Boston: Brill.

The position of Mardin Arabic

189

Jastrow, Otto. 2006b (appeared 2010). “Arabic dialects in Turkey – towards a comparative typology“. In: Türk Dilleri Arastırmaları 16: 153-164. Istanbul. Jastrow, Otto. 2006–2007. “Where do we stand in the research on the Anatolian qǝltu dialects?“ In: Nadia Anghelescu (ed.). Peripheral Arabic Dialects (= Romano-Arabica VI-VII), 63-69. Bucharest. Jastow, Otto 2013. “Grammaticalizations based on the verb kāna in Arabic dialects“. In: Clive Holes and Rudolf de Jong (eds.). Ingham of Arabia. A Collection of Articles presented as a Tribute to the Career of Bruce Ingham, 109-118. Leiden/Boston: Brill. Talay, Shabo. 2001. “Der arabische Dialekt von Hasköy (Dēr Khāṣ), Ostanatolien. I. Grammatische Skizze.“ In: Zeitschrift für arabische Linguistik 40: 71-89. Talay, Shabo. 2002. “Der arabische Dialekt von Hasköy (Dēr Khāṣ), Ostanatolien. II. Texte und Glossar“. In: Zeitschrift für arabische Linguistik 41: 46-86. Talay, Shabo. 2012. “Anmerkungen zum Vokalsystem des arabischen Dialekts von Sine, Provinz Diyarbakir, Osttürkei.“ In: Folia Orientalia 49: 545-555. Talay, Shabo. to appear. “The Mesopotamian-Levantine Dialect Continuum.“ Wittrich, Michaela. 2001. Der arabische Dialekt von Āzǝx (= Semitica Viva 25). Wiesbaden: Harrassowitz.

Tillo. Two texts reflecting daily life and cultural aspects of the Arabs of Tillo, South-eastern Turkey Ablahad Lahdo, Uppsala University

1.

Introduction

Between the years 2000 and 2003 and while working on my Ph.D. thesis, The Arabic Dialect of Tillo in the Region of Siirt (South-eastern Turkey), I conducted several fieldwork campaigns in Turkey. During my sole visit to Tillo in October 2000 I tape-recorded one of the two texts below, namely text 2. Text 1 was taperecorded by the informant himself and sent to me in March 2001. Due to volume difficulties I could not include these texts in my thesis and hence I find this opportunity suitable to publish them hoping that my esteemed colleague Professor Jan Retsö will appreciate them. The informant of these texts was at that time a 21-year-old man who studied to become a mullah. Despite his young age he was an excellent narrator and an articulate speaker of the dialect. Text 1 relates a day in the life of a young man. What is conspicuous in this text is that despite being a Sunni Muslim studying to become a mullah, and belonging to the Shāfiʿi school, the informant did not seem to have problems listening to music or watching television. This contrasts markedly with the attitudes and actions of certain radical young men within Islam today. Text 2 narrates a story of wisdom in an imaginary kingdom. 2.

Text 1

1. әl-lawm ǧītu mәn Baṭmāne. әl-ʿaraba ana k-aʿammәla. ka-maʿi ġayri arbaḥt әnfūs. aḷḷāhu aʿlam ṭarīq dǟm qaḍar sēʿa, bayn sēʿa sēʿa w nәṣṣ. ǧīna fәṭ-ṭarīq әl-ʿaskriyye saknūna. ʾēlu: kәs-saytәn әḥlēl fī ḥәqūq әt-trafīq.1 2. әšqat ʾәltulәn mǟ kәs-saytu ʾāl: bale kәs-sayt. katabūli ǧaza. baʿd ạ̄q ǧītu dawәmtu fī ṭarīqi lәḥәqtu l-bayt. lәḥәqtu l-bayt ḥaṭaytu l-ʿaraba šā Saʿīt ābe.2 uwwe rā әl-walēye ana bәqītu fī Tәllo.                                                                                                                         1 Cf. Turkish trafic ‘traffic’.

Two texts reflecting daily life and cultural aspects of the Arabs of Tillo

191  

3. yānlәẓ3 ka-kәl-bәqītu, šayš kәntu fī Baṭmāne ka-yǧīlna msāfrīn, bowš dostīn ka-kәlbәqītu bala nawm. әltu d-anamli mәqdār šī mā ṣār nasīp. ǧā wēḥәt arqaḍāš, әbәn ʿammo ḥaǧǧi Aḥmēt, ʾāl: taʿa tәndōrna4 dawra šī! әtūlu:5 xa-ysīr. 4. ǧā rәkәzna. ʾәlna, ltūlu: yālnәẓ ham nәrkaz šaʿrūke šī. rәkәzna. akalnālna šī mme. ṭalaʿna rәḥna lal-baqqāl. ṭabi l-baqqāl awne fi-yu ʿәndu anwēʿ bowš ammēn. axavnēlna mәnnu ham pasta6 w axavnēlna čuqqullāṭa w axavna qōla. 5. nēlu7 ham nәtmaš šaʿrūke šī. ṭabi ana mō ysīrli fәrṣa lay d-atmašša, dawāmli8 šayš ana fәl-muṭālaʿa. ana nāqәṣ ḥarake yaʿni. әtmaššayna mәqdārәn wēḥәt. lәḥәqna, fī tappatәn9 wәḥde fī Tәllo, ǧbēl yaʿne, fī yʾәlūlu quḅḅat әl-Xaṣya. ysēr mәn maymētna iyye amme, lәḥәqna ạwnak. 6. fataḥna ạw l-malzamāt lay b-īdna, akalnēhәn w šәrәbna mašrūbātna mme. ʾәlna tanәtmaš10 šaʿrūkat lәxxe. әtmaššayna. faraqtu fī rōḥi kәl-baqa aḅṭal. ǧīna hayyēne baytәn, fәn-nazle, ʾәltūlu: ana d-arkēp ʿal-faraṣ. 7. ʾālli: ana mme d-arkēp. ṭabi uwwe maww bōš ṣuwāri. ana mme mā ana bowš mali ama ham ana aḥsēn mәnnu. ḥaṭṭayna ʿalayu zžīn w әl-mәrġēp әmme ḥaṭṭaynēhu fī fәmmu. u dīn ana rәkәbtu ʿalayu. mәn zamēn šayš aḥat mā kes-sāq әl-faraṣ әyšayṭan bowš. 8. sәqtuhu kam karra. baʿәt lay rtaxa uwwe am rakkәptūhu. baʿәt lay rәkēp, zaǧǧaynālna ēke kam ṭūr 11 ʾәlna dә-nrōḥ l-bayt gērin. 12 ǧīna tәnrәtt fәl-maǧwe arayna wēḥәd arqadāš13 taḥtu ʿaraba ysōqa w yәǧi surʿat-li.14 mәn ǧamp әl-lәxxe әl-kalb baqa yәʿdi xalf әl-ʿaraba.

                                                                                                                                                                                                                                                                                                                                                            2 Cf. Turkish abi ‘elder brother’. 3 Cf. Turkish yalnız ‘but, only’. 4 Cf. tәndorlna in dativus ethicus ‘we make us a walk, tour’. 5 Cf. ʾәltūlu ‘I said to him’. 6 Cf. Turkish pasta ‘cake’. 7 Cf. ʾәlnēlu ‘we said to him’. 8 Cf. Turkish devamlı ‘continuous, lasting, unbroken’. 9 Cf. Turkish tepe ‘hill, mound’. 10 Cf. nәtmašša ‘we walk’. 11 Cf. Turkish tur ‘trip, excursion’. 12 Cf. Kurdish gerîn ‘stroll, move around’. 13 Cf. Turkish arkadaş ‘friend’. 14 Cf. Turkish -li which makes adjectives of substantives.

192

Ablahad Lahdo

9. ǧā ēke mәqdār ka.., ka-d-nardaysa15 ka-tyәṣṭaḥ16 fīna әl-kalp. ṭabi nǧarrayna ʿala ǧanp mәšš xāṭar mā yәṣṭaḥ fīna. rafēqna m sakkәn, ǟv әl-arqadāš lay әš-šüför, lay yʿammәl l-ʿaraba, sakkәn. rәkәbna fī ʿarabәtu. ǧīna hayyāne b-bayt. waqt lay rәkәbtu fәl-ʿaraba ka-kәl-naʿʿēm ḥәss taybu.17 ka-yġanni ġәnēni bōš bōš kwayyәs fәt-tayp. 10. ʾәltulu: xayr uww ǟv, әšš kayf-lәk әl-lawm ēke tġanni ạw әl-ġәnēni yġannaw fī tayb ēke? ʾalli: waḷḷa mәš xāṭәrak kәl-naʿʿamtu18 ḥәssu aṃa әṣṭәnat! fateḥ ḥәssu ǟk ṭallaʿna mә.., kǟnu bәl-kәrdi amme. kǟnu ġnēni ṭaybīn әnne. ǧīna hayyēne b-bayt. fәl-bayt nazzәlna. karra ǧīna l-bayt ḥaṿṿәrna, ṣārәt qarīp әl-ʿaša, ḥaṿru ʿaša faqaṭ ṣṭanṿәrna Saʿīt ābe hayyәne s-sēʿa fmēnye, fmēnye w nәṣṣ. 11. Saʿīt ābe ǧā. kәǧǧā ḥamūhu amme mәn ḥačč. ʾәlnēlu: yәlzam nrōḥ lә-zyārәtu. uwwe am arāha mwāfqa. wāldәti am arata mwāfqa. akalna akәlna. ʾәlna tәnsayy, tәnrōḥ әlwalēye. qәmna rәḥna l-walēye mәšš xāṭәr әl-ḥačč. әl-ḥaǧǧi tәnbērәklu mәšš xāṭar ḥaǧǧu. 12. rәḥna fәṭ-ṭaqṣi. lәḥәqna dǟm baytēn. ṭalaʿna fәd-dәrēyāč. әnne mā-lәn aṣanṣōr19 ṭalaʿna fәd-dәrēyāč. daqqayna z-zīl.20 fataḥūlna l-bēp. karmūna l-ōḍa. daxalna l-ōḍa rәkәzna ʿal-qultġāt. 21 ǧābū-lna fәl-bīr 22 ṃay zamzēm 23 w gābū-lna qasp. 24 baʿәd ạ̄k ǧābū-lna čēye, aḷḷa yәʿmar baytēn. ana әl-čēye ašraba fәl-fәnǧēn әz-zġīr. šrәbtūli qaḍar xams, sәtt fәnēǧīn, aḷḷāhu aʿalm. 13. ạk karra ka-kә-nfataḥ әt-talafәzyōn. ka-fi-yu fәlīm ạw lal-ʿayṭāt. 25 kәllәtna baqa nәtfarrač ʿat-talafәzyōn. әṣ-ṣәḥbe nqaṭaʿәt. baʿәt lay tamm ạ̄k ǧābū-lna mayva26 mme. kәǧ-ǧābu, fәl-mayva kǟnu ṭamāṭ.., kǟnu pәrtaqān, ka-fī tәffēḥ, ka-fī xyār qatt lay yәǧi mbēli. 14. akalna. faqaṭ ana kaltu pәrtaqāne b-waḥda. ǧābūlna, ṭabi fәs-sәkēkīn faṣṣaynāhәn. tammayna. rafaʿuwan ạwlak әmme. ṭallaʿna z-zamēn kәl-ʿaḅar ʿalayu bowš. ṣārat                                                                                                                         15 Cf. Turkish nerede ise ‘before long, pretty soon’. 16 To hit s.o. by accident. 17 Cf. Turkish teyp ‘tape-recorder’. 18 Cf. The Syriac root nʿm, neʿmoṯo ‘gentle sound or voice, soft whisper’. 19 Cf. Turkish asansör ‘elevator’. 20 Cf. Turkish zil ‘bell’. 21 Cf. Turkish koltuk ‘armchair’. 22 Cf. Turkish bir ‘one, here: at first’. 23 Cf. zamzam ‘well in mecca and that contains holy water’. 24 Cf. Kurdish qesp ‘dates’. 25 Kung fu movie. 26 Cf. Turkish meyve, meyva ‘fruit’.

Two texts reflecting daily life and cultural aspects of the Arabs of Tillo

193  

qarǧīp sāʿa fnaḥš rәdna tәnqūm. ṭalabna musāʿada mәnnәn. nәzәlna kәləәtna sayyna maʿәn wada. 15. ǧīna fī ṭarīq Tәllo; ṭabi kә-ṿlam әl-hawa. fәl-ṿalām amme s-seyēḥa iyy bōš bōš ṭaybe. lәḥәqna l-bayt. aḷḷāhu aʿlam ka-kәl-ʿaḅarat әs-sēʿa fnaḥš әnn ka-mā na ġalaṭ. ka-fī maʿna msēfrīn әmme. ạwlak әm waddaynēhәn, xallaynēhәn fī baytәn w ǧīna. 16. Saʿīt ābe ʾēl: ǟf ʿAbdullʾaḥḥat yәlzam әntәm šәġlu. ē dīn ana ʾәltu d-agri šaʿrūke šī mәšš xāṭar ǟv әl-qāsēt әmme. inšāʾa llāh kәṣ-ṣār šī әkwayyәs. ṭabi nәrǧi mәn aḷḷa ǧalla ǧalāluhu әntәn amme tәtәtʿawlna. 3. Translation text 1 1. Today I came from Batman. I was driving the car. Besides myself, I had four other people with me. The trip took, God knows, about one hour, between one hour and one hour and a half. While on the way the military stopped us. They said: you broke the traffic rules. 2. No matter how much I said that I did not he said: yes you did. They wrote me out a fine. After that I continued my way and arrived at home. I came home and gave the car to my older brother, Sait. He went to the town and I stayed in Tillo. 3. Since I was in Batman, guests and many friends came to visit us; so I was in need of sleep. I thought that I will sleep for a while but did not have a chance. A friend of mine, the son of uncle Haj Ahmet came to me and said: come on, let us have a walk! I said: so be it. 4. He came in and sat down. We said, I said to him: let us sit down for a while. We sat down. We also had something to eat. Then we went out to the grocer’s store. The grocer’s store here has, of course, a large variety of things. We bought from the grocer’s store pastry, choclate and cola. 5. I said to my friend: let us walk for a while. Normally I do not have time to walk because I always spend my time reading, so I lack exercise. We strolled for a while. We came to, there is a hill in Tillo, a mountain that is called the dome of Xasya, a decendant from our great-grandmother, we arrived there. 6. We unpacked what we had in our hands. We ate it and we also drank our drinks. We thought of walking a little bit more. We walked. I notice that I was getting tired. When we reached his house, down the hill, I said: I will ride the horse.

194

Ablahad Lahdo

7. He said to me: I too want to ride. He is, of course not very good at riding. I am not very good at riding either but still I am better than him. We saddled the horse and we also put the bridle on his muzzle. So I got on the horse. But because nobody has ridden the horse for a long time he fooled around a lot. 8. I rode the horse to and fro a couple of times. When the horse calmed down I took my friend also on the horse. When he also had ridden we made a couple of tours and then we thought of strolling home. On our way back we met a friend of ours driving fast in a car. On the other side (of the road) a dog started to run after the car. 9. The dog came about, it was just about to collide with us. Of course we stepped aside to avoid the collision. Our friend also stopped, the driver, the one who was driving the car, he stopped. We stepped into his car. We came all the way home. When I stepped into the car he had turned down the music. The songs in the tape-recorder were very very good. 10. I said to him: what is the occasion today for you to play these songs in your tape-recorder? He answered: I turned it down for your sake but listen! He turned it up and we noticed that the singing was in Kurdish. The songs were nice. We arrived at home, where he let us stepp out of the car. We came home and we prepared, it was about dinner-time, they prepared dinner but we had to wait for Sait, until eight o‘clock, eight thirty. 11. Sait arrived. It seemed that his father in law had come back from the pilgrimage. We said to Said: we have to pay him a visit. Sait also found that suitable, and in the same way my mother also found that suitable. We had our dinner. Then we though of.., we left to the town. So we left to the town for the sake of pilgrimage. We have to congratulate the Haj for his pilgrimage. 12. We went there by car. We arrived by their house. We took the stairs up. They do not have an elevator so we took the stairs up. We rang the door bell. They opened the door for us. They invited us into the living room. We entered the living room and sat on the armchairs. At first they offered us water (from the holy well of Zamzam) and dates. After that they brought us tea, may God keep their house safe. I usually drink tea in small cups. God knows, I drank about five, six cups. 13. Then, the television was switched on. A movie of screaming (kung fu) was on the television. We all started watching television then. We stopped talking with each other. When that was finished they offered us fruit. They brought, the fruit consisted of toma.., there was orange, apple and cucumbers, as I remember.

Two texts reflecting daily life and cultural aspects of the Arabs of Tillo

195  

14. We ate. I ate only one orange. They brought us, we used of course a knife to peel the fruit. We finished that. They cleaned those things. We noticed that the time passed and it became late. It was about twelve o’clock and we wanted to leave. We asked their permission for that. We took our leave and went down. 15. We drove on the way to Tillo; of course it became dark. It is very very nice to travel in the dark. We arrived at home. If I am not mistaken, the time had passed twelve o’clock. We also had some guests with us. We drove those and dropped them at their place and we came home. 16. Sait said: we have to finish Ablahad’s work. So I thought that I will taperecord a little to fill this cassette. I hope that it became something useful. We pray to God the Sublime that you also pray for us. 4. Text 2 1. yā ʿAbdәllaḥḥat, nәḥe awne fә-mamlakәtna27 šā axx әl-gbīr yʾulūlu ābe. ṭabi ǟvi lkәlme mәn luġat әt-tәrki kәl-ʿaḅarat ʿәnәdna, ana mme aʾәllәk ābe allāhu aʿlam. s-saḥ fī ʿәndi ktēp әysayy baḥs ḥәkēkin wēḥәt. mәn ạwl әl-ḥәkēki wәḥde mme d-aḥti.., d-aḥkī-kәn iyye inšāʾa llāh. 2. әyʾūl: yowm mәn әl-әyyēm fī fә-mamlakәtәn wәḥde pādišāh. ǟv әl-pādišāh kәṣ-ṣār bowš zaki, bowš ʿạ̄qạl, ʿēdәl. u ʿayni28 zamān-da kәṣ-ṣārlu wazīr әmme ham ʿayni šakәlda kәṣ-ṣār bōš bōš zaki. 3. ǟvi l-mamlake kәṣ-ṣār fī nәṣṣa bīrәn wәḥde. aṃa ṃayya kәṣ-ṣār bōš ḥәlu. ǧamīʿ mәllәt.., ǧamīʿ әl-mәlle mәn ǟv әl-bīr yәšrabu ṃayy čünkü29 fī ạ̄k әl-mamlake mā kәṣ-ṣār kamēha ṃayy ṭayyap. 4. yawm mәn әl-әyyēm kәr-rāḥ mnaǧǧamәn wēḥәt ạwnak kәl-ʾallәn: ǟvi l-bīr lā tәbdaw tәšrabu mәnna ṃayy! lay yәšrap mәnna ṃayy yā tәymūt yā tәyǧәn. ṭabi mō yṣadqu. lay mō yṣadqu šayy ǟv әẓ-ẓalame yәbd.., ysawaw dawām ʿala šәrp әl-ṃayy. 5. yәšrabu zämǟdәd wēḥәt. mәn yәšrap mәnnәn әyǧәnn aṃa mawt mōy fī. aksar nēs әyǧәnnu. ḥatta yәʿḅar ēke zamēn fәl-ṿayʿa ġayr әl-wazīr u ġayr әl-qәrāl,30 pādišāh šī aḥat mō yәbqa ʿạ̄qạl. 6. yawm әl-pādišāh әyʾūl: yāhu aṃa d-anzēl bayn ạwl mәllәti aṭallēḥ әšš ysawaw әšš mō ysawaw. ayy sәp yәtmašš kәlla yʾәlūlu: maǧnūn. ṭabi ạwl ḥarakāt әl-ʿạ̄qạl ʿәnt әl                                                                                                                         27 Cf. Turkish memleket ‘home district, country’. 28 Cf. Turkish aynı ‘the same, identical’. 29 Cf. Turkish ҫünkü ‘also, too’. 30 Cf. Turkish kral ‘king’.

196

Ablahad Lahdo

maǧnūn әysīr kama l-mәǧēnīn. čünkü әnne fī nafsu rōḥan yṭәššu rōḥan ʿala ṭabiʿat әlḥaqq. 7. baʿat lal-pādišāh yәfraq kәlla mәllәtu kәǧ-ǧannēt uwwe b-waḥdu ʿạ̄qạl mō tәyәġbәlu ʿalayu, mō tәyxallawu fī rāsan, yʾūl šā wazīru: ḥāl ǧībūli mәn ạ̄k әl-ṃayy! yәʾmәr šā lwazīr. әl-wazīr yәšrap u baʿәt әl-pādišāh amm, әl-qәrāl amm yәšrap. әfnaynәtәn amm әyǧәnnu. 8. ạ̄k әč-čēx31 l-mamlake kәlla, ašxāṣ lay fәl-mamlake әšš әrǧēl, әšš nәswēn, әšš awlēt, kәlla yʾәlu: hēš uww pādišāhna, hēš kurālna, hēš mal.., malikna. ǧā ʿala ṣirāṭ әl-ḥaqq. 5. Translation text 2 1. Ablahad, here in our home district one names the older brother as Abe. This word has of course come to us from Turkish, nevertheless I too, by all means will call you Abe. Now, I have a book that contains stories. From these stories I will tell you one, if God wills. 2. It says: once upon a time there was a sultan in a kingdom. This sultan was very clever, very wise and just. At the same time he had a minister who also was very very clever. 3. In the middle of this kingdom there was a well. The water of this well was very sweet. All his peop.., the whole population drank from this well because in that kingdom there wasn’t water as delicious as that. 4. One day an astrologer came there and said to them: stop drinking water from this well! He who drinks water from it, either he will die or he will go mad. But of course they did not believe that. These who did not believe the astrologer went on drinking the water. 5. They drank for some time. Those of them who drank went mad but there was no death. The majority of people went mad. When some time passed, in the whole village, no one except the minister and the sultan, the king kept their sanity. 6. One day the sultan thinks: maybe I should get out and stroll among my subjects and see what they are up to. Wherever he walks every one shouts at him and says: crazy. Of course the crazy ones see the sane one’s behavior as if he were crazy. Because the crazy people consider themselves as normal.

                                                                                                                        31 Cf. Kurdish ҫax ‘when, at what time’.

Two texts reflecting daily life and cultural aspects of the Arabs of Tillo

197  

7. When the sultan noticed that all his subjects had gone crazy and that he was the only one that stayed normal, he feared that they wouldn‘t accept him anymore, they wouldn‘t keep him as their sultan. He said to his minister: bring me some of that water immediately! He gave on order to the minister. The minister drank and thereafter the sultan also, the king drank. Also the two of them went crazy. 8. Then the whole kingdom, the subjects in the kingdom, men, women and children said: he is still our sultan, still our king, still our kin.., our king. He found the path of righteousness.

Une occurrence ancienne de la structure kāna sa-yafʿalu en arabe écrit * Pierre Larcher, Aix-Marseille Université, Aix-en-Provence What is Arabic ? Jan Retsö (2013)

Abstract In Modern Standard Arabic one encounters a structure kāna sa-yafʿalu equivalent to the French « il allait faire » or to the English « he was going to do » and having like those, among other possible interpretations, that of the past conditional (« il aurait fait » / « he would have done »). A similar structure exists in the dialects. One might therefore think of an innovation – influenced by the dialects – of Modern Standard Arabic, what might be called a “classicised dialectalism”. Unfortunately, an example of kāna sayafʿalu appears in Sībawayhi’s (d. 179/795?) Kitāb concerning the description of law and thus seeming to confirm the primordial character of its interpretation as past conditional. Rather than segmenting Arabic grammar into a grammar of Classical Arabic and one of Modern Standard Arabic, it would be wiser to build a historical grammar of written Arabic.

Résumé On rencontre en arabe moderne une structure kāna sa-yafʿalu, équivalent au français « il allait faire » ou à l’anglais « he was going to do » et ayant, comme eux, entre autres interprétations possibles, celle de conditionnel passé (« il aurait fait » / « he would have done »). Une structure analogue existe dans les dialectes. On pourrait donc penser à une innovation de l’arabe moderne, influencée par les dialectes, ce qu’on pourrait appeler un « dialectalisme classicisé ». Malheureusement, un exemple de kāna sa-yafʿalu apparaît dans le Kitāb de Sībawayhi (m. 179/795 ?), à propos de la description de law, semblant ainsi confirmer le caractère primordial de l’interprétation comme conditionnel passé. Plutôt que de segmenter la grammaire de l’arabe en grammaire de l’arabe classique et grammaire de l’arabe moderne, il serait plus judicieux de construire une grammaire historique de l’arabe écrit.

*

Merci à Catherine Pinon pour m’avoir fourni les références exactes des exemples (9)–(13).

Une occurrence ancienne de la structure kāna sa-yafʿalu

199

1. Introduction Ceux des arabisants qui, pour des raisons pédagogiques et/ou de recherche, sont amenés à lire la presse arabe y rencontrent occasionnellement une structure que je noterai kāna sa-yafʿalu : il s’agit en effet de sa-yafʿalu, c’est-à-dire de l’inaccompli indicatif du verbe lié à la particule dite du futur sa-, mais placé dans le champ du verbe opérateur kāna à l’accompli. Si, comme l’auteur de ces lignes, ils ont un certain âge, ils ne se souviennent pas de l’avoir jamais rencontrée au temps lointain de leurs études (fin des années 60, début des années 70 du siècle dernier) et, sauf erreur de ma part, elle ne figure pas dans les grammaires de l’arabe classique. On est donc tenté de conclure qu’il s’agit d’une innovation de l’arabe moderne. Pour ma part, je me souviens exactement de la première fois où je l’ai rencontrée. C’était dans un article du journal tunisien al-Ṣabāḥ du 14/12/2001 1 : (1) ḫāṣṣatan ʾanna ʾiqnāʿahu bi-ʾinfāq ʾamwālihi fī bilādihi mubādara yumkin ʾan tuwaffir li-l-bank al-markazī kammiyya hāʾila min al-ʿumla al-ṣaʿba kānat satuḥawwal ʾilā ḫāriǧ Tūnis « Surtout que le [i.e. le Tunisien] convaincre de dépenser son argent dans son pays peut procurer à la banque centrale une quantité formidable de devises, qui allait être transférée [= aurait été transférée] hors de Tunisie ». Je me souviens aussi de ne pas l’avoir immédiatement comprise et d’avoir pensé d’abord à une espèce de calque, peut-être (s’agissant d’un journal tunisien) du français. En effet, en arabe moderne, sa-yafʿalu tend à s’imposer à côté de faʿala et yafʿalu, comme un troisième « temps », faisant de yafʿalu un pur « présent », alors qu’en arabe classique, yafʿalu s’oppose à faʿala, sur le plan temporel, comme un non-passé à un passé, laissant au contexte l’interprétation de ce non-passé comme présent ou comme futur. C’est d’ailleurs parce que, selon la formule consacrée, yafʿalu « se dit équivoquement du présent et du futur » (muštarak bayna al-ḥāl wa-l-mustaqbal) que sa- est dit « rendre yafʿalu purement futur » (sa- yuḫaliṣṣuhu li-l-istiqbāl) 2. Si l’on pense que sa-yafʿalu est l’équivalent de « il fera », mettre ce sa-yafʿalu dans le champ de kāna, qui, dans un emploi libre, est essentiellement un exposant temporel du passé, peut sembler l’ana1

Cet exemple est cité dans Larcher (2006 et 2007), mais sa troncation entraîne une mauvaise lecture de yumkin ʾan tuwaffir en yumakkin ʾan tuwaffar…

2

Ibn Hišām al-ʾAnṣārī (m. 760/1361), Muġnī al-labīb, I, 147.

200

Pierre Larcher

logue du conditionnel présent (« futur dans le passé ») des langues indoeuropéennes. Si l’on y réfléchit, on verra que celui-ci, morphologiquement, n’est jamais que l’adjonction des marques du passé à des formes qui, sans elles, sont celles du futur… En français, futur simple et conditionnel présent sont construits sur le même radical, qui est en fait l’infinitif du verbe. Si l’on compare leurs paradigmes respectifs, on voit nettement que le conditionnel présent ajoute à ce radical les mêmes désinences que l’imparfait (temporellement passé et aspectuellement imperfectif) ajoute au radical simple (il chanter-ait / il chantait). On voit aussi, aux 1re et 2e personnes du pluriel, que si on ôte le i, on a le futur simple (nous chanter-i-ons / chanter-ons ; vous chanter-i-ez / chanter-ez) et que si l’on ôte le suffixe –er de l’infinitif, on a le présent (nous chant-er-ons / chant-ons ; vous chant-er-ez / chant-ez). On retrouve, mutatis mutandis, la même proportion dans une langue comme l’anglais ; le futur y est périphrastique, composé du présent d’un auxiliaire de mode (shall/will) et de l’infinitif, mais c’est la substitution au présent du passé (should/would) de cet auxiliaire qui donne le conditionnel présent. J’étais cependant conscient qu’une interprétation comme conditionnel présent ne faisait pas sens dans le contexte. Par bonheur, je rencontrais bientôt une autre occurrence de la même structure, de peu antérieure, dans le journal égyptien al-ʾAhrām du 5/7/2000, ce qui montrait qu’elle n’était pas liée à une région particulière du monde arabe. Et, cette fois-ci, elle figurait dans un contexte qui permettait de l’interpréter correctement : (2) law kānat ʾilīzābīṯ qad ʾanfaqat niṣf al-waqt allaḏī qaḍat-hu fī riʿāyat ǧiyādihi ʿalā al-ihtimām bi-ʾabnāʾi-hā fa-ʾinna-hā kānat sa-tuǧannib al-ʾusra almuškilāt... « si Elisabeth avait consacré la moitié du temps qu’elle a passé à élever ses chevaux à l’éducation de ses fils, elle aurait évité à sa famille les problèmes » 3. Law est en effet la marque de l’irréel. Mais alors qu’en arabe classique law faʿala est ambigu, pouvant, selon le contexte, s’interpréter soit comme irréel du présent (« s’il faisait ») soit comme irréel du passé (« s’il avait fait »), en arabe moderne law kāna faʿala se désigne aussitôt comme irréel du passé : en plaçant faʿala dans le champ de kāna, on restaure sa valeur aspectuelle d’accompli qui 3

Exemple cité dans Larcher (2006).

Une occurrence ancienne de la structure kāna sa-yafʿalu

201

est à l’inverse neutralisée quand il est placé directement dans le champ de law. Comme marque d’une hypothèse irréalisable, law constitue généralement la protase d’un système hypothétique, attendant une apodose. La liaison de l’apodose à la protase a considérablement évolué au fil du temps. En arabe classique, c’est-à-dire tel qu’enseigné dans les classes, elle est la-faʿala à la forme positive et mā faʿala à la forme négative. S’il est vrai qu’en arabe coranique la- ne manque presque jamais, en arabe médiéval il manque souvent, de sorte que, la forme positive de l’apodose étant simplement faʿala, la forme négative devient lam yafʿal. Il arrive aussi que, mā faʿala n’étant plus compris comme la négation de la-faʿala, il est précédé de la- et qu’on trouve la-mā faʿala, qu’on ne trouve en revanche jamais en arabe coranique. Il y a une bonne raison à cela : mā faʿala est la négation de faʿala dans les dialectes. On peut alors dire que la-mā faʿala s’apparente à une hypercorrection. Toujours en arabe médiéval, on trouve aussi l’apodose yafʿalu, qui tire l’interprétation du système hypothétique law p, q, où p et q sont des propositions, du côté de l’irréel du présent. Enfin l’arabe moderne renoue avec la segmentation, au sens du linguiste suisse Charles Bally (1865– 1947) (Bally 1965), de l’arabe ancien, mais d’une autre manière, en substituant au segmentateur la- le segmentateur fa-. Le statut de segmentateur de la- est bien établi par le fait qu’il se retrouve, sous le nom de lām al-taʾkīd (« lām de corroboration ») et non plus de lām al-ǧawāb (« lām d’apodose »), dans une phrase telle que ʾinna Zaydan la-qāʾimun (« oui, Zayd, il est bel et bien debout ») : il y délimite le thème ʾinna Zaydan, probablement issu d’une ancienne phrase (quelque chose comme « voici Zayd »), du propos qāʾimun (« [il est] debout ») 4. Dans la mesure où kānat sa-tuǧannib figure ici dans l’apodose d’un système hypothétique en law, que sa protase désigne comme irréel du passé, il n’y a pas d’autre interprétation possible que « elle aurait évité ». Au demeurant, les étudiants arabophones de mon cours, qui ont pour langue maternelle un dialecte arabe, non seulement me confirmaient cette interprétation, mais encore m’indiquaient qu’ils ont dans leurs parlers l’analogue de cette structure : on peut en effet placer dans le champ de kān le futur périphrastique de ces parlers, composé de l’inaccompli lié à une particule (e.g. syrien rāḥ yākol « il va manger ») 5.

4

Ce développement sur law reprend Larcher (2003).

5

Ce que j’ai dû savoir dans une vie antérieure. L’auteur de ces lignes a successivement appris et oublié trois dialectes arabes (syrien, libyen oriental, marocain), le premier moins que les deux autres, du fait d’un second séjour, plus récent, en

202

Pierre Larcher

J’étais donc tenté de conclure non seulement à une innovation de l’arabe moderne, mais encore à une innovation apparue sous l’influence des dialectes, autrement dit un dialectalisme « classicisé ». 2. L’occurrence Dans ce contexte, la thèse de Giolfo (2010 : 42) me remit en mémoire une occurrence de kāna sa-yafʿalu que j’avais oubliée, mais que j’avais toutes les raisons personnelles de connaître : elle figure en effet dans un article de Versteegh (1991 : 79) publié dans une livraison thématique du Bulletin d’Études Orientales que j’avais dirigée lors de mon séjour à l’Institut Français d’Études Arabes de Damas (IFEAD) en 1989-1990. Elle est extraite du Kitāb (IV, 224) de Sībawayhi (m. 179/795 ?) : (3) wa-ʾammā law fa-li-mā kāna sa-yaqaʿu li-wuqūʿ ġayri-hi. L’intérêt de cette occurrence, c’est qu’elle concerne le sens de law, qui a ici le statut d’autonyme. Et même si Sībawayhi est fort peu prolixe sur law (comparativement à ʾin) et ne le décrit pas encore, comme le feront ses successeurs, en termes de « condition » (šarṭ) d’une part et d’ « impossibilité » (imtināʿ) d’autre part, l’emploi même de kāna sa-yafʿalu le désigne bien comme un irréel du passé. Giolfo (2010 : 42) traduit d’ailleurs (3) en français par : (3a) « law est pour ce qui serait arrivé si quelque chose d’autre était arrivé » et Giolfo (2012 : 155) le traduit en anglais par : (3b) « law is for what could have happened if something else had happened », transformant en subordonnée conditionnelle, avec plus-que-parfait, ce qui en arabe même est un syntagme prépositionnel de cause (li-), où se trouve enchâssé un syntagme nominal ayant la structure d’une iḍāfa, le premier N[om] étant un maṣdar (neutre du point de vue du temps/aspect et, quand le verbe est transitif, du point de vue de la voix), soit :

Syrie. La situation politique du monde arabe n’en facilite pas la pratique quotidienne…

Une occurrence ancienne de la structure kāna sa-yafʿalu

203

(3c) « Quant à law, il est utilisé pour ce qui serait arrivé, du fait de l’arrivée d’une autre chose ». Le contexte exclut donc d’emblée une autre interprétation de kāna sa-yafʿalu à laquelle on aurait pu penser, du fait de son apparition dans une relative : non plus un emploi libre, où kāna a sa valeur temporelle de passé (ou aspectuelle d’accompli), mais un emploi lié à celui d’une particule qui le détermine et le neutralise (on pourrait parler d’un kāna purement syntaxique et non plus sémantique). Cet emploi se rencontre en arabe moderne et je l’y avais d’ailleurs rencontré dès avant le kāna sa-yafʿalu qui nous intéresse ici, par exemple dans le journal jordanien al-Šaʿb du 23/4/1994 : (4) Wa-lammā kāna Krīstūfir wazīr al-ḫāriǧiyya al-ʾamrīkī sa-yabdaʾ bi-ǧawla šarq ʾawsaṭiyya fī al-sāʿāt al-qalīla al-muqbila ʿalā ʿamal tanšīṭ al-ʿamaliyya al-silmiyya fa-ʾinna qaḍiyyat al-ḥiṣār yaǧib ʾan taḥtall ḥayyizan kabīran fī dāʾirat ihtimāmātihi... « Et vu que (étant donné que, attendu que, puisque) le secrétaire d’état américain Christopher va entamer une tournée moyen-orientale dans les toutes prochaines heures, dans l’espoir d’activer le processus de paix, la question du blocus occupera nécessairement une grande place dans le cercle de ses préoccupations » 6. Nous avons bien ici une occurrence de sa-yafʿalu dans le champ de kāna et, sur le plan strictement syntaxique, il n’y a aucune différence entre cette occurrence et celles déjà rencontrées : dans les termes de la grammaire arabe traditionnelle, un verbe opérateur appliqué à une phrase nominale, ayant pour thème (ism kāna) Christopher et pour propos (ḫabar kāna) la phrase verbale sa-yabdaʾ…. Mais sur le plan sémantique, il y a une considérable différence : kāna est ici déterminé par lammā. Au cours de l’histoire, lammā a considérablement évolué, tant sur le plan sémantique que syntaxique, mais au cours de cette histoire, il n’en a pas moins conservé deux traits : 1) lammā détermine faʿala ; 2) lammā faʿala sert de cadre à l’énonciation d’une autre phrase. Mais alors qu’en arabe classique, lammā faʿala est le thème d’un propos étant lui-même une phrase verbale à l’accompli, avec lequel il constitue une phrase (faiblement) segmentée lammā faʿala faʿala de sens temporel (« lorsqu’il fit…, il fit… »), en arabe 6

Exemple cité dans Larcher (2003 : 281, n. 13).

204

Pierre Larcher

moderne 7, il a pris un sens justificatif et constitue avec son propos une phrase (fortement) segmentée au moyen de fa-, à l’instar de beaucoup des complexes de phrases de l’arabe moderne (conditionnels, concessifs etc.) 8. Cependant, Versteegh (1991 : 79) est plus littéral, qui traduit (3) par : (3d) « law is used (…) for what which was going to happen on account of the occurrence of something else ». La traduction « littérale » de Versteegh a, à mes yeux, un mérite. On a vu en introduction que rencontrant pour la première fois la structure kāna sa-yafʿalu j’avais hésité, pensant d’abord à une interprétation comme conditionnel présent, suggérée par le fait que la structure se présente comme un « futur dans le passé », avant de me rallier, du fait du contexte, à une interprétation comme conditionnel passé. Mais si l’hésitation est une vertu en matière de recherche, elle est un vice en matière de pédagogie. Aussi pour faire comprendre aux étudiants francophones qui, en tant que tels, devaient avoir autant de mal que moi à comprendre cette structure immédiatement, avais-je trouvé un « truc » : sayafʿalu restant analysable, il faut le comparer, non au futur « il fera », mais au futur périphrastique, encore appelé proche (mais la première appellation, purement descriptive, est préférable à la seconde qui impose une interprétation), « il va faire ». Celui-ci est composé de l’indicatif présent du verbe aller et de l’infinitif du verbe. Son correspondant passé est il allait faire. « Il allait faire » correspond bien à l’anglais he was going to do, lui-même correspondant passé de he is going to do. Or, comme je le notais (Larcher, 2006 : 58, n. 5 ; 2007 : 88, n. 17), le français il allait faire a au moins un emploi où il a une interprétation comme irréel du passé (= « il aurait fait ») : c’est quand il est couplé à une subordonnée circonstancielle avec un verbe au passé, ce que j’exemplifiais par : (5) J’allais te le dire, quand tu m’as interrompu, qui est logiquement l’équivalent de : (6) Je te l’aurais dit, si tu ne m’avais pas interrompu. 7

En fait le lammā justificatif se rencontre déjà dans la prose technique médiévale.

8

Sur les « complexes de phrases », cf. Larcher (2008).

Une occurrence ancienne de la structure kāna sa-yafʿalu

205

En effet, les principales de (5) et (6) présupposent toutes deux « je ne te l’ai pas dit » et la subordonnée circonstancielle de (5) pose ce que la subordonnée conditionnelle de (6) présuppose, à savoir « tu m’as interrompu » (et qui est le contraire de ce qu’elle suppose à titre irréel). Elles sont donc bien logiquement équivalentes, le fait que tu m’aies interrompu ayant eu pour conséquence que je n’ai pu te le dire. On peut d’ailleurs, en français même, croiser les deux structures et dire (en marquant une pause forte entre les deux segments) : (7) J’allais te le dire… si tu ne m’avais pas interrompu ! C’est bien par référence à cette équivalence logique que je parlais d’interprétation « contrefactuelle » de il allait faire 9. La paraphrase de l’arabe kāna sa-yafʿalu par le français « il allait faire » avait ainsi plusieurs avantages. Sur le plan morphologique, une superposition des deux formes est possible : elles combinent toutes deux, même si ce n’est pas de la même façon, les trois morphèmes de temps passé (le verbe opérateur kāna en arabe, le suffixe -ait en français), de temps futur (la particule sa- en arabe, le verbe aller en français) et d’aspect inaccompli (la forme yafʿalu en arabe, l’infinitif simple du verbe en français). En revanche une telle superposition est impossible avec le conditionnel passé du français, qui, comme forme composée, marque l’aspect accompli, qui, en arabe, est marqué par faʿala. Sur le plan des emplois, il en a bien un où il est l’équivalent du conditionnel passé. Mais, comme je le notais, il allait faire a d’autres emplois en français et notamment un que j’appelais contrastivement « factuel » et dont je donnais pour exemple (Larcher 2006 : 58, n. 5 ; Larcher 2007 : 88, n. 17) : (8) Karol Wojtyla fut élu pape en 1978. Sous le nom de Jean-Paul II, il allait régner près de vingt-sept ans. En fait, cet emploi « factuel » (futur dans le passé, le passé étant ici historique) n’est lui-même qu’un des emplois possibles de il allait faire en français. Il allait

9

On peut d’ailleurs supposer que Versteegh choisissait la traduction « littérale » was going to happen parce que, dans le contexte, elle laissait la porte ouverte à l’interprétation could have happened. Manuela Giolfo (communication personnelle) me confirme qu’on peut dire en anglais I was going to call you, when you arrived avec le sens de I could have called you, if you hadn’t arrived.

206

Pierre Larcher

faire étant à il va faire ce que il ferait est à il fera, on peut le trouver à la place de il ferait, là où il va faire peut se trouver à la place de il fera. On a donc toute une gamme d’emplois de il allait faire, allant du plus factuel au plus fictif. La comparaison de l’arabe kāna sa-yafʿalu avec le français il allait faire avait ainsi, last but not least, un dernier avantage, d’entrouvrir, par ricochet, la porte à une question : l’arabe kāna sa-yafʿalu a-t-il tout ou partie des emplois du français il allait faire, dont certains sont aussi ceux de l’anglais he was going to do ? Pour ma part, n’ayant rencontré, au gré de mes lectures (des articles de journaux destinés à mon cours d’arabe de presse), que très peu d’occurrences de kāna sa-yafʿalu, toutes « contrefactuelles », j’inclinais à penser qu’à la différence du français il allait faire il ne semblait pas avoir d’autre interprétation… Par la porte ainsi entrouverte s’engouffre alors Pinon (2012) dans sa thèse consacrée à ce qu’elle appelle joliment « la nébuleuse de kān » et fondée sur un large corpus, tout à la fois diatopique et multigenres d’arabe moderne (presse, mais aussi littérature et même blogs). Elle en extrait 99 occurrences de kāna sayafʿalu (et analogues dialectaux également présents dans son corpus, qui fait ainsi place non seulement à la variation géographique, mais encore linguistique). 48 correspondent à ce que j’ai appelé ci-dessus des emplois liés, où sayafʿalu est dans le champ de kāna, lui-même lié à une particule qui le détermine : ils ne nous intéressent pas ici. Restent donc 39 emplois libres + 12 dialectaux. Si l’on regarde les quatre exemples qu’elle donne (p. 272-273) comme représentatifs de ces 39 occurrences, on s’aperçoit qu’un seul est « contrefactuel », au sens où nous l’entendons, c’est-à-dire devant être traduit en français par un conditionnel passé (journal tunisien al-Ṣaḥāfa du 13/04/2011) 10 : (9) kāna al-baḥr hādiʾan ṭiwāl al-ʾarbaʿ wa-l-ʿišrīn sāʿa al-māḍiya wa-ʾillā fa-ʾinna alwaḍʿ kāna sa-yataḥawwal ʾilā kāriṯa « La mer était calme tout au long des dernières vingt-quatre heures : sinon, la situation se serait transformée en castastrophe ». On note tout de suite qu’il s’agit en fait, comme (2), d’un emploi « corrélatif », kāna sa-yafʿalu figurant dans l’apodose d’un système hypothétique dont la protase est ʾillā : celle-ci, quoique formée sur ʾin, sert de pro-protase aussi bien aux 10 Titre de l’article : al-ḥaras al-baḥrī yunqiḏ maǧmūʿa min al-muhāǧirīn ġayr al-šarʿiyyīn (« Les garde-côtes sauvent un groupe d’immigrés illégaux »). En ligne sur : http://www.essahafa.info.tn.

Une occurrence ancienne de la structure kāna sa-yafʿalu

207

systèmes en ʾiḏā qu’en law. La protase serait ici en law (law lam yakun al-baḥr hādiʾan = « si la mer n’avait pas été calme »). Notons que l’absence de protase explicite n’empêche pas une interprétation « contrefactuelle ». Il suffit comme en (1) qu’elle soit implicite (« une quantité formidable de devises qui aurait été transférée hors du pays [si le Tunisien y avait passé ses vacances] »). Les trois autres exemples ne sont pas « contrefactuels », sans pourtant être uniformément « factuels », allant au contraire du plus factuel au plus fictif. Le premier (Balaḥ al-qurā « La datte des villages », nouvelle de Ǧubayr al-Mlīḥān, Arabie Saoudite, du 12/03/2005) 11 est : (10) ʾīh ʾayyuhā al-ǧāsūs qul man ʾarsalaka ? wa-mā allaḏī kunta sa-tafʿalu-hu fī hāḏihi al-qarya al-ʾāmina ? « Hé, l’espion ! Dis : qui t’a envoyé ? Et qu’allais-tu faire dans ce paisible village ? ». Il est factuel (puisque celui auquel on s’adresse est là !). Il s’agit bien d’un « futur dans le passé ». Le locuteur prend pour repère, non le moment où il parle (en ce cas, il aurait dit « que viens-tu faire dans ce paisible village ? »), mais celui, antérieur, où il imagine que quelqu’un a envoyé pour espionner celui qu’il interpelle. Le second (blog Vampitra, Maroc, du 11/08/2008)12 est : (11) qabl al-šahr kānat wafāt al-masīrī (…) allaḏī kuntu sa-ʾaktubu tadwīna ʾan wafātihi lākinnanī lā ʾataḏakkar mā allaḏī ḥālanī dūna ḏālika « Il y a un mois, survint la mort d’al-Masîrî, dont j’allais écrire la nécrologie (…), mais je ne me souviens pas de ce qui m’en a empêché ». Il n’est pas factuel, puisqu’il présuppose bien « je ne l’ai pas écrite », sans être exactement « contrefactuel », ne pouvant pas être paraphrasé par « je l’aurais écrite ». En revanche, il est paraphrasable, par « j’étais sur le point d’écrire », révélant qu’il s’agit d’un « futur proche dans le passé ». C’est en fait cet emploi qui fait le lien avec l’emploi « contrefactuel » (il a le même présupposé) : il suffirait d’ailleurs de réécrire « mais je ne me souviens pas de ce qui m’en a em11 Mise en ligne le 6/4/2005 sur : http://www.arabicstory.net/index.php?p=text&tid=5856. 12 En ligne sur : http://vamprita.wordpress.com.

208

Pierre Larcher

pêché » en « quand quelque chose, dont je ne me souviens pas, m’en a empêché » pour avoir cette interprétation. Le troisième (al-Mīrāṯ « L’héritage », nouvelle de Basma al-Šawwālī, Tunisie, du 6/8/2010) 13 est : (12) kānat sa-taǧhar bi-l-ġināʾ wa-tanfaǧir bi-ḍaḥk wa-kānat al-ʾaḥzāb almutaqābila sa-tattaḥid fī ḥaddi-hā bi-l-raǧm ǧazāʾ al-surūr badal min al-bukāʾ « elle chanterait en public et exploserait de rire 14, et les partis adverses s’uniraient pour la punir de la lapidation, rétribution de la joie, au lieu des larmes, [qu’elle aurait donnée] ». Il est purement fictif. Dans un récit au passé, un personnage, à un moment donné se projette en imagination dans l’avenir. Kāna sa-yafʿalu peut être traduit non seulement par il allait faire, mais encore il ferait, ce qui montre bien qu’une comparaison avec le conditionnel présent n’était pas spécieuse. Des 12 occurrences dialectales qu’elle donne en annexe (p. 524-525), seules six sont pertinentes 15 : quatre sont des apodoses de systèmes hypothétiques en law, deux des futurs proches dans le passé, liés implicitement ou explicitement à une circonstancielle au passé, comme dans (Hīlāna « Hélène », roman de Walīd al-Ḥaǧǧār, Syrie, 2001)16 :

13 En ligne sur : http://www.aladabia.net/ar/article-4751-3_1/. 14 On peut admettre que du fait de la coordination wa- et du parallélisme de sa construction avec celui qui précède (un syntagme prépositionnel en bi-), cet inaccompli est lui-même dans le champ de sa-. Une interprétation comme complément d’état (« en explosant de rire ») rendrait soit la coordination inutile, soit d’avoir en plus un pronom personnel (wa-hiya). 15 Ne le sont pas les occurrences de ʿam yafʿal dans le champ de kān, mais celles-ci ont le mérite de rappeler que « il était en train de faire », interprétation de cette structure, est une des deux interprétations possibles de he was doing, l’autre étant justement ... « il allait faire ». De la même façon, en effet, que le présent progressif peut marquer un futur proche (My grandmother is arriving tonight), le passé progressif peut marquer ce même futur après verbe au passé (I told her, grandmother was arriving that day). Les exemples sont tirés de Ogée et Boucher (1990 : 77 et 80). 16 En ligne sur http://www.walidalhajjar.com/pres/06.htm ou encore : http://syrianstory.com/amis-3-8.htm.

Une occurrence ancienne de la structure kāna sa-yafʿalu

209

(13) kān raḥ ymūt lammā šāf Bītū ʿam yǝḥtǝreʾ « Il allait mourir, quand Bîtû l’a vu en train de brûler ». Très pertinemment, Pinon renvoie à Kassab (1970 : 150), où figurent deux exemples de ce type, remontant donc à près d’un demi-siècle. La structure paraît ainsi dès longtemps installée dans les dialectes, spécialement dans l’emploi de « futur proche dans le passé », souvent lié à une circonstancielle, emploi conduisant à son apparition dans les apodoses des systèmes hypothétiques en law. Ce qui, par contrecoup, suggérerait que si les emplois du type de (9) et de (11) sont plutôt influencés par les dialectes, les emplois du type de (10) et (12) le sont plutôt par les langues européennes. Il est donc clair que kāna sa-yafʿalu mérite, en arabe moderne, l’enquête approfondie que Pinon se propose de mener. Enquête qui permettra de décrire et classer ses différents emplois d’une part, de faire éventuellement apparaître le rôle que peuvent y jouer les dialectes arabes et les langues européennes (anglais, français) d’autre part. 3. Conclusion L’apparition de kāna sa-yafʿalu avec le sens d’irréel du passé dans un texte de la fin du IIe/VIIIe siècle pose cependant un redoutable problème de linguistique historique : peut-on maintenir que kāna sa-yafʿalu est une innovation de l’arabe moderne, influencée par les dialectes ? Ou bien, une telle hypothèse ne résultet-elle pas du fait qu’entre les grammaires de l’arabe dit classique et les grammaires de l’arabe dit moderne, nous n’avons en fait rien, c’est-à-dire pas de grammaire historique de l’arabe ? La grammaire de l’arabe dit classique est un artefact. Elle repose sur un corpus qui, pour l’essentiel, est chronologiquement préclassique et elle ignore, non seulement tout ce qui vient après, mais encore ne tire pas toutes les conséquences de traits pourtant attestés dans ledit corpus. On enseigne ainsi que dans les systèmes potentiels ʾin p, q, où p et q sont des propositions, on trouve soit l’inaccompli apocopé yafʿal, soit l’accompli faʿala, la négation étant la même : lam yafʿal. L’examen du corpus coranique montre pourtant que yafʿal l’emporte très largement sur faʿala et que la négation de yafʿal est systématiquement lā yafʿal, lam yafʿal étant celle de faʿala qu’on trouve dans les systèmes éventuels en ʾiḏā : s’il y a quelques ʾin faʿala, faʿala dans le Coran, on n’y trouve encore aucun exemple ni de ʾin lam yafʿal, faʿala ni de ʾin faʿala, lam yafʿal. Le soi-disant

210

Pierre Larcher

système potentiel « classique » apparaît ainsi comme l’entrecroisement des deux systèmes potentiel et éventuel de l’arabe préclassique 17. La grammaire de l’arabe dit moderne n’est pas moins un artefact. Elle est née de la prise de conscience progressive des arabisants que l’évolution de l’arabe ne se réduisait pas à la modernisation du lexique, mais touchait aussi la grammaire. Cette prise de conscience ne remonte pas au delà de la deuxième moitié du XXe siècle : on se souvient par exemple de l’ouvrage pionnier de Monteil (1960). Les corpus, sur lesquels se fondent les grammaires de l’arabe dit moderne, étant contemporains de leurs auteurs, ils ne remontent eux-mêmes pas au delà. Par suite, ces grammaires désigneront comme « moderne » un trait comme celui que l’on rencontre dans l’exemple (4), qui ne concerne pas la syntaxe, mais la formation lexicale : l’adjectif de relation šarq ʾawsaṭ-ī (« moyenoriental »), formé sur le syntagme nominal mawṣūf/ṣifa (« objet qualifié/qualification ») al-šarq al-ʾawsaṭ (« Le Moyen Orient »). En arabe classique, on considère qu’en ce cas un adjectif de relation n’est formable qu’après « télescopage » (naḥt) des éléments de la base complexe en une forme faʿlal-, i.e. ʾImruʾ al-Qays > marqasī, ʿAbd Allāh > ʿabdalī etc. L’ennui, c’est que ladite « innovation » se rencontre à date ancienne, par exemple chez le géographe Muqaddasī (m. fin IVe/Xe siècle) disant d’Untel qu’« il était châfiʿite [pour l’école juridique] et ʾabû ʿamrawî [pour la lecture] » (kāna šafʿawiyyan ʾabū ʿamriyyan) 18, c’est-à-dire suivait dans la récitation du Coran la lecture de ʾAbū ʿAmr ibn ʿAlāʾ (m. 154/770). Il est facile de comprendre ce qui s’est passé : Muqaddasī est trop tardif pour la grammaire de l’arabe dit classique, il est trop ancien pour celle de l’arabe dit moderne. Même si près de deux siècles séparent Sībawayhi et Muqaddasī, ils n’en appartiennent pas moins tous les deux à ce même entredeux. Celui-ci, et ce n’est pas le moindre des paradoxes, est constamment parcouru par les arabisants, mais, sauf exceptions, pour son contenu, non la forme linguistique de ce contenu. Il est pourtant clair que Sībawayhi écrit une autre langue que celle qu’il décrit 19 et si un kāna sa-yafʿalu vient sous sa plume (le 17 Ce développement résume Larcher (2009). Hermann Reckendorf (1863–1924) avait remarqué que lā yafʿal était « à comprendre comme négation d’un apocopé » (lā mit Apok., das als Verneinung eines Apok. zu verstehen ist ») (Reckendorf 1921 : 487). 18 ʾAḥsan al-Taqāsīm, p. 202, l. 4, de l’édition De Goeje cité par Fück (1955 [1950] : 172). 19 Qui, par exemple, pourrait soutenir sans rire que le ʾiʿrāb y joue le moindre rôle, alors qu’il est central dans sa description de la ʿarabiyya ?

Une occurrence ancienne de la structure kāna sa-yafʿalu

211

Kitāb étant le livre d’un auteur), c’est qu’il était formable et interprétable. Ce qui rend peu probable qu’il s’agisse d’un hapax 20, quelque hypothèse que l’on formule sur son apparition (y compris celle que l’auteur ici n’est pas un arabophone natif). Au fond, variation linguistique et contact de langues sont les deux mêmes facteurs à prendre en considération, hier comme aujourd’hui. Il faut donc, comme disent les cyclistes, changer de braquet et passer à une grammaire historique de l’arabe écrit. Cette dernière étiquette a, à mes yeux, un avantage : d’inclure de facto le moyen arabe, dont l’étude est en plein développement, mais qui se concentre sur la « mixité » (le Middle Arabic étant devenu définitivement Mixed Arabic) créée dans la synchronie d’un même texte par la variation linguistique, alors que celle-ci en diachronie n’apparaît que comme un facteur, parmi d’autres, de l’évolution.

20 De fait, ce n’en est pas un, ne serait-ce que par le biais de la citation. Versteegh (1991 : 85) renvoie par exemple au Muqarrib (97.9-98.2) de Ibn ʿUṣfūr (m. 670/ 1271). On y trouve en effet par deux fois (p. 90 de notre édition) la définition de Sībawayhi, ce dernier non nommé. Ibn ʿUṣfūr y distingue deux law, « celui qui est une particule employée pour ce qui allait arriver du fait de l’arrivée d’autre chose ou qui a le sens de ʾin » (law allatī hiya ḥarf li-mā kāna sa-yaqaʿu li-wuqūʿ ġayrihi ʾaw bi-maʿnā ʾin). Il ajoute que « la différence entre les deux est que celui qui est pour ce qui allait arriver du fait de l’arrivée d’autre chose est suivi d’un verbe qui a le sens du passé, même s’il n’en a pas la forme (…) », confirmant ainsi l’interprétation de kāna sa-yafʿalu comme irréel du passé, tandis que « celui qui a le sens de ʾin rend le verbe purement futur : si sa forme est celle du passé, il en transforme le sens en futur » (wa-l-farq baynahumā ʾanna allatī hiya li-mā kāna sayaqaʿu li-wuqūʿ ġayrihi yakūnu al-fiʿl baʿdahu bi-maʿnā al-muḍiyy wa-ʾin lam takun ṣīġatuhu ṣīġat al-māḍī (…) wa-llatī hiya bi-maʿnā ʾin tuḫalliṣu al-fiʿl li-l-istiqbāl wa-ʾin kānat ṣīġatuhu ṣīġat al-māḍī ṣayyarat maʿnāhu li-l-istiqbāl). Ibn Ǧinnī (m. 392/1002) cite également (Ḫaṣāʾiṣ, III, p. 332) la structure kāna Zaydun sayaqūmu ʾamsi (« Zayd allait se lever hier »), avec le commentaire suivant : « c’est-à-dire il était attendu de lui qu’il se lève dans le temps passé » (ʾay kāna mutawaqqaʿan minhu al-qiyāmu fīmā maḍā).

212

Pierre Larcher

Bibliographie 1. Sources primaires Ibn Ǧinnī, Ḫaṣāʾiṣ = ʾAbū al-Fatḥ ʿUṯmān Ibn Ǧinnī, al-Ḫaṣāʾiṣ, éd. Muḥammad ʿAlī alNaǧǧār, 3 volumes. Beyrouth : Dār al-Hudā li-l-ṭibāʿa wa-l-našr. Ibn Hišām al-ʾAnṣārī, Muġnī al-labīb = Ǧamāl al-dīn Ibn Hišām al-ʾAnṣārī, Muġnī allabīb ʿan kutub al-ʾaʿārīb, éd. Māzin Mubārak, Muḥammad ʿAlī Ḥamd Allāh et Saʿīd al-ʾAfġānī, 2 volumes. Beyrouth : Dār al-fikr. 1969. Ibn ʿUṣfūr, Muqarrib = ʿAlī b. Muʾmin al-maʿrūf bi-Ibn ʿUṣfūr, al-Muqarrib, éd. ʾAḥmad ʿAbd al-Sattār al-Ǧawārī et ʿAbd Allāh al-Ǧubūrī. Bagdad : Maṭbaʿat alʾĀnī, 1971. Sībawayhi, Kitāb = Abū Bišr ʿAmr b. ʿUṯmān b. Qanbar Sībawayhi, al-Kitāb, éd. ʿAbd al-Salām Hārūn, 4 volumes. Beyrouth : ʿĀlam al-kutub. S.d.

2. Sources secondaires Bally, Charles (1965). Linguistique générale et linguistique française, 4e édition revue et corrigée. Berne : Francke. Fück, Johann (1955[1950]). ʿArabīya. Recherches sur l’histoire de la langue et du style arabe, traduction de Claude Denizeau, avec une préface de l’auteur et une introduction de Jean Cantineau [tr. fr. de ʿArabīya. Untersuchungen zur Sprach- und Stilgeschichte, Abhandlungen der sächsischen Akademie der Wissenschaften zu Leipzig, Philologisch-historische Klasse, Band 45, Heft 1. Berlin : Akademie Verlag]. Giolfo, Manuela (2010). Les systèmes hypothétiques en ʾin de l’arabe classique. Etude syntaxique et sémantique : une hypothèse modale. Thèse de doctorat. Aix-en-Provence : Université de Provence (Aix-Marseille I). Giolfo, Manuela (2012). « Yaqum vs qāma in the conditional context : a relavistic interpretation of the frontier between the prefixed and the suffixed conjugations of the Arabic language », dans The Foundations of Arabic Linguistics, Sībawayhi and Early Arabic Grammatical Theory, edited by Amal Elesha Marogy with a foreword by M.G. Carter. Brill : Leiden. Kassab, Jean (1970). Manuel du parler arabe moderne au Moyen-Orient, tome premier, Publications du Centre Universitaire des Langues Orientales vivantes, 6e série, tome VIII. Paris : Imprimerie Nationale et Librairie orientaliste Paul Geuthner. Larcher, Pierre (2003). « Les systèmes hypothétiques en law de l’arabe classique », Bulletin d’Études Orientales, t. LV (2003), p. 265-285. Damas : Institut Français d’Études Arabes. [http://halshs.archives-ouvertes.fr/halshs-00132072/en/] Larcher, Pierre (2006) « Le ‘segmentateur’ fa-(ʾinna) en arabe classique et moderne », Kervan. Rivista internazionale di studii afroasiatici/International Journal of Afro-Asiatic Studies, Gennaio 2006, numero 3 / January 2006, num. 3, p. 51-63. [http://www.cisi.unito.it/kervan/contents/documents/3_5_LAR.pdf] Larcher, Pierre (2007). « L’arabe classique : trop de négations pour qu’il n’y en ait pas quelques-unes de modales », dans Christian Touratier et Charles Zaremba (éds) La

Une occurrence ancienne de la structure kāna sa-yafʿalu

213

Négation, Travaux Linguistiques du CLAIX n° 20, p. 69-90. Aix-en-Provence : Publications de l’Université de Provence, 2007. Larcher, Pierre (2008). « Les ‘complexes de phrases’ de l’arabe classique », Kervan. Rivista internazionale di studii afroasiatici / International Journal of Afro-Asiatic Studies, Luglio 2008 n. 6 / July 2008 n. 6, p. 29-45. [http://www.kervan.unito.it/contents/documents/n_6_4_LARCHER.pdf] Larcher, Pierre (2009). « Les systèmes conditionnels en ʾin de l’arabe classique », Bulletin d’Études Orientales LVIII [2008, année de tomaison], p. 205-232. Damas : Institut Français du Proche-Orient. [http://www.cairn.info/resume.php?ID_ARTICLE=BEO_058_0205]. Monteil, Vincent (1960). L’arabe moderne. Paris : Klincksieck. Ogée, Frédéric et Boucher, Paul (1990). Grammaire appliquée de l’anglais. Paris: Sedes. Pinon, Catherine (2012). La nébuleuse de kān : classification des différents emplois de kānayakūnu à partir d’un corpus d’arabe contemporain. Thèse de doctorat. Aix-enProvence : Université d’Aix-Marseille. [http://www.theses.fr/2012AIXM3078/document]. Reckendorf, Hermann (1921). Arabische Syntax. Heidelberg: C. Winter [2e édition, 1977]. Retsö, Jan (2013). « What is Arabic ? », dans Jonathan Owens (éd.) The Oxford Handbook of Arabic Linguistics, p. 433-450. New York : Oxford University Press. Versteegh, Kees (1991). « Two conceptions of irreality in Arabic grammar », dans Pierre Larcher (éd) De la grammaire de l’arabe aux grammaires des arabes, Bulletin d’Études Orientales, tome XLIII, p. 77-92. Damas : Institut Français d’Études Arabes.

Luġat al-ʾumm and al-luġa al-ʾumm – the ’mother tongue’ in the Arabic context Gunvor Mejdell, University of Oslo

1.

Celebrating the International Mother Tongue Day

In Tawasol/Tawāṣul 10 (2009) – a bilingual publication issued by the Oman Commission for Education, Culture and Science, the editorial celebrates the UN International Year of Languages as well as the UNESCO International Mother Tongue Day on February 21st of every year. The English version of the editorial has the heading “The Mother Language”, the Arabic version is titled ”Al-luġa al-ʾumm”. It deplores the ”demise of mother tongues (al-luġāt al-ʾumm) across the world” as a consequence of globalisation, and supports the efforts made by international cultural institutions to counter these developments and preserve linguistic and, thus, cultural diversity. The editorial (both versions) cites the Cambridge Dictionary definition of mother tongue as ”the language spoken by the child while he is still in his infancy” (al-luġa allatī yataḥaddaṯu-hā al-ṭifl wa-huwa lā yazāl yaḥbū (< the child when still crawling)” and that ”whatever language the child learns after that is (to be considered) his second language”. However, the article does not hesitate to link directly to al-luġa al-ʿarabiyya (’the Arabic language’) – with its 1600 years of existence as a living language, including its role as the language of the Qurʾān, ”reaching wherever Islam reached”, and concludes: fa-ʾaṣbaḥat luġat alʿilm wa-l-ʾadab wa-l-siyāsa wa-l-ḥaḍāra faḍlan ʿan kawni-hā luġat al-dīn wa-l-ʿibāda (”[to] become the language of science and literature, politics and civilization as well as the language of religion and worship”). The editorial calls on ”all of us to ensure that the Arabic language, our mother tongue (al-luġa al-ʿarabiyya alʾumm), continues to develop in order to help secure our Arab culture (al-ʾamn al-ṯaqāfī al-ʿarabī)”. One observes in the text a glide from the notion of ’mother tongue’ as a child´s first language towards a notion of it as a language of religion and high culture. These notions are normally separated in socio-linguistic literature, applying terms such as ’vernacular’ versus ’standard language’, ’Low variety’

Luġat al-ʾumm and al-luġa al-ʾumm

215

versus ’High variety’ (in diglossia), in Arabic terms al-ʿāmmiyya versus (alʿarabiyya) al-fuṣḥā. The commonly understood use of the term1 ’mother tongue’ is reflected in the dictionary definition cited in the editorial.2 The UNESCO definition from its ’foundational’ 1953 report is also quite explicit: ”the language which a person acquires in early years and which normally becomes his natural instrument of thought and communication”.3 All speakers of Arabic know that their everyday spoken language is of another kind than the standard variety they learn to read and write in school, and they have a label for it, ʿāmmiyya or dārija (or other local variants). In international education, including the UNESCO programmes, the call for using the mother tongue/vernacular in primary education is a campaign for training literacy via the child´s first language (or something close to it), as a pedagogical principle, i.e. that building on what the child is acquainted with, facilitates his/her access to writing and reading and other cognitive skills. Extending the label ’mother tongue’ to cover the language of high culture, religion, and literature as we saw in the Tawāṣul editorial, can be considered, from a language ideology perspective, as an instance of ’erasure’: “Erasure is the process by which ideology, in simplifying the sociolinguistic field, renders some persons or activities (or sociolinguistic phenomena) invisible. Facts that are inconsistent with the ideological scheme go unnoticed, or are explained away. For example, a social group or language may be imagined as homogeneous, its internal variation disregarded” (Gal 1998). In this case it is the distance between the first language of the Arab child and the standard ʿarabiyya, the situation of diglossia, which is erased from the ’narrative’. This is typical of official discourse by representatives of what John Eisele has labelled ”the dominant regime of authority” on Arabic linguistic issues,4 whose language

1

’Mother tongue’ itself is not really a technical term in the socio-linguistic tradition(s), apart from studies on language policy and educational issues, especially in multilingual contexts.

2

Similarly: ”the language which a person has grown up speaking from early childhood” (online Oxford Dictionaries) also ’native language’, ’first language’).

3

Cited in Fasold 1997, 246. The report is: The Use of Vernacular Languages in Education. Paris: UNESCO, 1953.

4

Eisele 2003. For a discussion, see Mejdell 2008.

216

Gunvor Mejdell

ideology insists on the continuity, unity, and purity of the Arabic language, and as far as possible neglects the role of the spoken vernaculars.5 While the Tawāṣul editorial simply passes over what seems to be inherently contradictory, the following article (in Arabic), actually confronts the issue of language acquisition and linguistic socialisation of children. In ”al-ʿarabiyya ... al-luġa al-ʾumm wa-ʾaṯaru-hā fī takwīn al-ṭifl” (Al-ʿarabiyya – the mother tongue/language and its effect on the child´s development/formation), Dr. Yaḥyā Farġal ʿAbd al-Muḥsin of the Department of Arabic studies, Sultan Qābūs University, argues: ”Standard Arabic (al-arabiyya al-faṣīḥa) is the mother tongue/language for all Arabs who have adopted it as official language for spoken and written communication. We do not agree on the definition of mother tongue with those who see it as the dialect which the child acquires (al-lahja allatī yaktasibu-hā al-ṭifl) – by that they mean the vernacular dialects (yaqṣidūna bi-ḏālik al-lahajāt al-ʿāmmiyya) – arguing that it is the first to find its way to the child´s ear (ʾawwal mā yataṭarraq ʾilā masāmiʿhi)” (ʿAbd al-Muḥsin 2009: 22).

Then follows a highly rhetorical sequence on the natural disposition/propensity of Arabic-speaking people from early history for refined taste in language; how the revelation of Islam, besides making al-ʿarabiyya the language of faith and worship (ʿaqīda wa-taʿabbud), further strenghtened it in excellence and clarity, and strenghtened the grip of the language on the hearts of its speakers and listeners (ibid.: 23). The advent of the Qurʾān brought the Arabs together on the basis of the common high Arabic language, the mother tongue/language (alluġa al-ʿarabiyya al-fuṣḥā al-muštaraka – al-luġa al-ʾumm), and the language of the Qurʾān became the absolute ideal linguistic style. Thus – when children happen to hear reciters of the Qurʾān and start listening to them, they will acquire its eloquent style. They will first listen and imitate, and then gradually understand. ʿAbd al-Muḥsin then cites from the translated work of a specialist in child language development, Sergio Sabini (?), to the effect that by the age of five or six, the child will have the ability to master his mother tongue/language (luġata-hu al-ʾumm), in order to achieve his needs and goal. The child gradually expands his grammar, and by the age of six he has finished his basic linguistic training (qad ʾatamma tadrība-hu al-luġawī al-ʾasāsī)” (ibid.: 26). On the authority of a European educationalist, not writing on Arabic – the readers should think 5

They nowadays have a Facebook ’community’ (since Oct. 2012) with 804 ’likes’ (ʾiʿjābāt): al-luġa al-ʾumm al-luġa al-ʿarabiyya al-fuṣḥā. https://ar-ar.facebook. com/alarabiaalfsha (downloaded 19.06.14).

Luġat al-ʾumm and al-luġa al-ʾumm

217

the child at six speaks al-luġa al-ʿarabiyya al-faṣīḥa? We must conclude that some wishful thinking, or ideological erasing, is involved in this line of argument. At this stage it should be stated that no educated Arab, no teacher of Arabic will explicitly deny that no one speaks fuṣḥā as from the cradle (or even from the age of six), or deny that ʿāmmiyya is the idiom spoken at home. Still the fact is, ”that in the educational field [the fuṣḥā] is referred to as the mother language (al-luġa al-ʾumm)” (Suleiman 2013: 272). It definitely is the target norm of Arabic language education, and the ultimate aim of promoting the language is ”to help secure our Arabic culture”. Thus, the orientation and argument is not about promoting literacy and educational goals, but about preserving a cultural (and religious) tradition. On the other hand, but also in connection with the International Mother Tongue Day, professor Danḥā (?) Ṭūbiyā Kūrkīs of Jadāra university in Jordan, posts a text (21.02.10)6 criticising those Arabs who in the media promotes alʿarabiyya al-fuṣḥā on this occasion, which is not strictly speaking a mother tongue to any living Arabic speaker. By ’mother tongue’ the UN intends, he argues, ”those languages threatened by extinction, among them the modern Aramaic dialects (lahajāt) spoken spontaneously by real mothers, in contrast with those who claim that al-ʿarabiyya al-fuṣḥā is their mother tongue”. The professor proposes a distinction between ”luġat al-ʾumm – mother tongue” and ”al-luġa al-qawmiyya – native language”. The former is to be reserved for the first language acquired by a person, the latter for the majority language in the state in which the person lives. Sometimes the two will coincide, as in the case of a person growing up with an Arabic dialect (lahja dārija) as his mother tongue and al-ʿarabiyya al-fuṣḥā as his native language. Kūrkīs, however, speaks for the minority languages in the region, and without challenging the status of alʿarabiyya, calls for respect and a place for these in the educational system. The clash in perceptions of ’mother tongue’ and its ideological implications, have also been an issue in the field of ’mother tongue’ training for immigrant children, which was widely discussed (as well as practised) in several Western countries. In Scandinavia, modersmål/morsmålsundervisning was organised in order to promote the child´s cognitive skills through using and developing his/ her ’own’, innate, language – and even in order to better have access to the new language (Swedish/Norwegian in this case). The problem was that ’mother 6

http://ankawa.com/forum/index.php?PHPSESSID=u0d8nvf... 04.01.13)

(downloaded

218

Gunvor Mejdell

tongue’ was taken to be standard Arabic in the case of Arab children, (and even grotesquely in the case of Berber speaking children), as the parents wanted their children to learn the language which was important in their home (e.g., Moroccan) school system and which had prestige (while Norwegian educators were unaware of the problem, confident as they were in their experience with a ’mother tongue’ quite close to the national standard). The result was an even greater stress and pressure on the children, who had to cope with an extra language (variety).7 2. ’Mother tongue’ and the transfer of concepts ”Originally a Latin term, lingua materna was used in the Middle Ages and the Early Modern period to refer to people´s vernaculars in contrast to learned Latin” (Yildiz 2012: 10). The context thus is one of 'vernacularisation' where local Romance language varieties began to be written and gradually replaced Latin in several language domains. The English ‘mother tongue’ was – according to the online Oxford English Dictionary – attested as early as the 15th century (’modre-tongue’). From either of those sources ’mother tongue’ spread as a term (calque) across languages: Muttersprache (German), modersmål/morsmål (Scandinavian), mitriki glossa (Greek), mātṛ bhāṣā (Hindi), materština (Czech), anadil (Turkish), bogo (Japanese), muyu (Chinese) (but Russian: rodnoj jazyk = ‘native’ language). I have not been able to search into when exactly, or via what source, the concept was adopted into Arabic,8 but somehow it appeared in two parallel constructions: luġat al-ʾumm (‘the mother´s language’) and al-luġa alʾumm (the mother language) – to be discussed below. The metaphor of language as ’mother’, is a ”gendered and affectively charged kinship concept” with strong ”emotional and ideological connotations” (Yildiz 2012: 10); it is the language of the mother, of the family, of daily social life in contrast with the High language and culture of the Church and the cultural establishment (prototype: Latin). Thus one would expect the transfer of the concept into Arabic, into luġat al-ʾumm / al-luġa al-ʾumm to refer to the local Arabic vernacular, al-ʿāmmiyya, the variety which is associated with values and expressions of intimacy, in family or love relations. The High language, alʿarabiyya al-fuṣḥā, or simply al-fuṣḥā, is associated with authority, with religious 7

This was also the case for Pakistani children, mostly with a Punjabi speaking background, who were taught Urdu as their ’mother tongue’. Similar problems are documented in Davies and Bentahila 1989.

8

Suleiman 2013, writing about these notions, has no reference to early uses of it.

Luġat al-ʾumm and al-luġa al-ʾumm

219

(Islamic) and state and educational authority, easily with connotations of paternalistic authority. However, and simultaneously, it is also strongly imbued with aesthetic, cultural and religious values, considered uniquely beautiful and expressive among the languages of the world. Finally, it symbolises the power and unity of the community/nation, al-ʾumma, whether understood as Muslim or Arab. There are reasons to assume that the concept of ’mother tongue’, as so many other European concepts, was transferred to the Arab world during the later nahḍa period. At that time, the nationalist idea of the mother tongue/Muttersprache as the unified and unifying language of the people/nation was very influential, and the emerging Arab nationalists adopted ”the highly consequential political linkage of language and nation”. The European romantic nationalists (Yildiz specifically mentions the German thinkers Herder and Schleiermacher) ”celebrated the distinctness of each language […] emanating from the genius of a particular nation [and] insisted on the need to maintain the distinctness of these national languages lest they lose their authenticity and rootedness in their respective nations” (Yildiz 2012: 6f.). The pride of the Arab nationalists in their cultural and linguistic heritage, of which al-ʿarabiyya al-fuṣḥā was the vehicle, was central to the nahḍa – so, I suggest, it was the standard variety, not the vernacular Arabic, which for them was imbued with the symbolic role of the unifying and authentic ‘Muttersprache’. (And still is: A page on the website of the Baath party (from January 2011) bearing the title ”The national significance of the mother tongue” (al-ʾahammiyya al-qawmiyya li-l-luġa al-ʾumm) actually refers precisely to Herder´s views on the organic connection of the language with its people´s thought and traditions.9) As one discussant in my data argues (see below), al-ʿarabiyya should be labeled luġat al-ʾumma (the language of the community/nation) in contrast with luġat al-ʾumm (the language of the mother). The phonetic proximity and similarity of the two concepts ʾumm and ʾumma no doubt facilitates the perceptual link between them. It may also have played a role in transferring the European nationalist idea of the ’mother tongue’/’Muttersprache’ linked to the people/ nation/sons: the luġat al-ʾumm as the embodiment of the ʾumma – whether of the Muslim community or the Arab nation of the nahḍa in the late 19th century, both with ’high’ Arabic as the language of its culture and civilisation. 9

http://www. baath-party.org/index.php?option=com_content&v... (downloaded 19.06.14)

220

Gunvor Mejdell

3. luġat al-ʾumm versus al-luġa al-ʾumm One finds much controversy over the form as well as content of the ’mother tongue’ equivalents in Arabic. I shall in the following present some examples from diverse sources. In the article ”The relationship between ”translation” and ”the mother tongue” (al-ʿalāqa bayna ”al-tarjama” wa-”al-luġa al-ʾumm”) Māhir (2009) raises the issue of how to understand the expression al-luġa al-ʾumm. Compared to the European languages from which the expression was imported, ”we have changed the meaning from their luġat al-ʾumm and tend to use it as luġat alʾumma (the language of the community/nation)”. To the Egyptians, Egypt (miṣr) is their mother (ʾumm) – consequently luġat miṣr would be the natural luġat al-ʾumm or al-luġa al-ʾumm, the language of dreams and hopes. While the refined language of classical Arabism (al-ʿurūba al-turāṯiyya) is the language of all the Arabs, on the national and official level, superposed to all colloquial varieties, the language of writing, commanded by grammar and the dictionaries and preserved in the Qurʾān etc. However, writes Mahir, it is not the language that the child learns spontaneously from his parents. While Māhir thus admits that ’mother tongue’ refers to something else in the Arabic context than in Europe, he does not distinguish between the two forms, or constructions, that are used in Arabic for ’mother tongue’. On the Egyptian (maṣri) Wikipedia,10 al-luġa al-ʾumm is defined as: ”the first language the child takes from his parents, that is, he doesn’t have to study it in order to speak it” (hiyya ʾawwal luġa b-yāḫud-ha il-ṭifl min ʾahl-u, yaʿni miš biyiḥtāg yudrus-ha ʿalašān yitkallim-ha). There is, however, a small, but significant, difference between the term al-luġa al-ʾumm and al-luġa al-rasmiyya: in spite of the fact that the Arabic language (al-luġa al-ʿarabi) is the official language of Egypt, it is not the mother tongue/language of the Egyptians (miš hiyya al-luġa al-ʾumm bitaʿt il-maṣriyyīn).” The entry then refers to the situation of diglossia (al-izdiwāgiyya al-luġawiyya) in the Middle East and North Africa, as well as other communities in which the language of writing is other than their mother tongue/language.11 10 http://arz.wikipedia.org.wiki/‫ة األم‬# ‫غ‬# ‫ل‬# # # # # # ‫( ال‬downloaded 19.06.14). I transcribe the technical terms in their standard variants, otherwise according to an Egyptian Arabic reading. 11 ‫لغه‬#‫عربى هى ال‬#‫لغه ال‬#‫ن إن ال‬#‫م م‬#‫رغ‬#‫ يعنى على ال‬.‫مّيه‬#‫رس‬#‫لغه ال‬#‫لغة اُالم و ال‬#‫ني مصطلح ال‬#‫ ب‬,‫هم‬#‫س م‬#‫ ب‬,‫سيط‬#‫رق ب‬#‫يه ف‬#‫ف‬ ‫وده فى‬# # ‫وج‬# # ‫يا( م‬# # ‫لوس‬# # ‫ة )ديج‬# # ‫لغوي‬# # ‫ّية ال‬# # ‫ اإلزدواج‬.‫ني‬# #‫صري‬# #‫ة امل‬# # ‫تاع‬# # ‫لغه االم ب‬# # ‫ش هى ال‬# # ‫كن م‬# # ‫ ل‬,‫صر‬# # ‫مّيه فى م‬# # ‫رس‬# # ‫ال‬

Luġat al-ʾumm and al-luġa al-ʾumm

221

Searching the web for entries on ‫ ل " " " " " "غ"ة األم‬one finds a wide range of entries discussing this concept. In an exchange (ʾijābāt Google)12 the question posed February 17, 2011 is simply: mā hiya luġat al-ʾumm? – ”awamia” responds: what is intended by al-luġa al-ʾumm is the original language (al-luġa al-ʾaṣliyya) spoken by people, thus: al-luġa al-ʾumm for the Saudis and the Egyptians = al-ʿarabiyya; al-luġa al-ʾumm for the United States = alinglīziyya; – ”Narges” responds: al-luġa al-ʾumm is the first language the person learns from childhood (ʾawwal luġa yataʿallamu-hā al-ʾinsān munḏu ṭufūlati-hi) – ”umm rūdīnā”, ”love love”, ”hasanalsheikh”, ”mahi” and ”hajar” all respond: (al-luġa) al-ʿarabiyya, and ”umm rūdīnā” adds: wa-l-luġa al-ʿarabiyya hiya al-luġa allatī ʾūnzila [sic] bihā l-qurʾān (the Arabic language is the language in which the Qurʾān was revealed); On another Google exchange 13 the question goes: li-māḏā summiyat alluġa al-ʿarabiyya bi-l-luġa al-ʾumm? (why was the Arabic language called the mother tongue/language?) – the ’most favoured’ response argues that it is because al-luġa al-ʿarabiyya is considered the Semitic language which is closest to the al-luġa al-sāmīya alʾumm, the Semitic Ursprache; – another response takes the issue back to the Creation of Adam and the first human language; the notion ʿarab goes back to the descendants of the lineage from rab (‫( )رب‬two others lineages were mar (‫ )م " " " " " "ر‬and sar (‫ – )س " " " " " "ر‬all meaning sayyid/master). Then biradicals became triradicals, (in short) through addition of ʾalif > ʾarab, amor(ites) and ʾasar (i.e. suryān) – then there was a shift from ʾalif to ʿayn > ʿarab, the other clans left north and west, while the ʿarab remained in the jazīra. More straightforward are the responses: – li-ʾanna-hā luġat al-qurʾān al-karīm (because it is the language of the Qurʾān), and – li-ʾannahā ʾasmā wa-ʾarqā luġāt al-ʿālam (because it is the loftiest and finest language in the world; ‫غتها اُالم فى‬# ‫ير ل‬# ‫ه غ‬# ‫كتوب‬# ‫غه م‬# ‫تعمل ل‬# ‫ه بتس‬# ‫قيا و أى مجتمع فى دول‬# ‫ري‬# ‫مال اف‬# ‫ط و ش‬# ‫األوس‬-‫رق‬# ‫جتمعات دول الش‬# ‫م‬ ‫الكتابه‬. 12 http://ejabat.google.com/ejabat/thread?tid=55cc8d8c372f4193 19.06.14) 13 http://ejabat.google.com/ejabat/thread?tid=740cbe80d6ac0a0d

(downloaded

222

Gunvor Mejdell

– yimkin li-ʾanna-hā il-luġa il-waḥīda illi taḥtawi ʿala ḥarf al-ḍād (‫[ )ظ‬sic] (maybe because it is the only language with the letter/sound ḍād (‫)ظ‬ or even cooler: – wa-man allaḏī sammāhā bi-ḏālik? (and who was it that called it so?) On Jan. 25, 2009, on a site for language and translation issues, ”Nesrin” posts the following comment and question:14 al-luġa al-ʾumm...hākaḏā naqūl bi-l-ʿarabiyya wa-lākin suʾāl ʿindī ḥawl hāḏā huwa limāḏā laysa luġat al-ʾumm ḥayṯu yakūn tarkību-hā muḍāf wa-muḍāf ʾilay-hi – ʾiḏan, mā ʾiʿrāb al-luġa al-ʾumm? (we say al-luġa al-ʾumm in Arabic, but I have a question about that: why isn’t it luġat al-ʾumm, with the genitive/ʾiḍāfa construction, so, what is the ’grammatical reason’ for al-luġa al-ʾumm?) Her question received 13 comments, and 25 votes are given to the preferred variant: – 14 in favour of al-luġa al-ʾumm (said to denote the most widespread (al-ʾakṯar šuyūʿan) language in a country, and that luġat al-ʾumm ”gives another meaning”; – 6 are in favour of luġat al-ʾumm (not specifying what it denotes), while – 5 respondents support the entry which argues that both forms are correct. This entry is by Sumaya from Jordan, who explains: Both are correct, but there is a slight difference (farq basīṭ): when we say al-luġa al-ʾumm (grammatically mawṣūf wa-ṣifa (approx. head and qualifier), syntactically mubtadaʾ wa-ḫabar (subject and predicate [sic]) ”the mother qualifies the language, and the intended meaning is that it is the first and basic language which the child learns from birth, while when we say luġat al-ʾumm we intend that the child at birth learns the language of his parents, and the language is associated with the mother, since the mother is the one who interacts more with the child, and one sees him learn most of his words from her. Thus the difference is negligable.” Sumaya thus interprets both forms in accordance with the original intension of ’mother tongue’. Judging from the citations in arabiCorpus of the string ‫( ل" " " " "غة األم‬for which the definite form with article, ‫ ال " " " " " "لغة‬, also displays) there appears, however, a certain distributional pattern, namely a tendency to use the ʾiḍāfa (genitive) construction luġat al-ʾumm in the sense of ’the language spoken by the (child’s) mother’, i.e. the ’mother’s tongue’ in the concrete, primary meaning, slightly extended to 14 http://www.proz.com/kudoz/Arabic/linguistics/3047731 (downloaded 19.06.14)

Luġat al-ʾumm and al-luġa al-ʾumm

223

convey the notion of ’first language’, ’native language’ of a person. The appositive construction ‫’ ال " " " " " "ل"غ"ة األم‬the language the mother’ > ’the mother language’ is more ambiguous: often the distinction seems to be blurred (as we saw in the Tawāṣul editorial) between the primary meaning of ’mother language’ (implicitly the spoken vernacular) and the extended cultural and social (and ideological) meaning as ’the community language of national culture and identity’ (implicitly the standard language) – or, sometimes even in the historical sense of an ’original language’, or Ursprache. a) all instances (4 of a total of 57 citations) of luġat al-ʾumm: Mother language ‫ لغة األم‬،‫بإمكان الطفل اكتساب لغة أخرى باإلضافة إلى اللغة األساسية‬ ”the child’s ability to acquire another language in addition to his basic language, (i.e.) luġat al-ʾumm Mother language” ‫ سواء اإلسبانية أو غيرها‬،‫الذي نشأ عن الزواج املختلط يتقن إلى جانب اللغة العربية لغة األم أيضا‬ ”the one who grew up in a mixed marriage, masters, besides Arabic, the mother’s language as well, be it Spanish or otherwise” ‫الصراع األول الذي عاشه بني لغة األم ولغة االب‬ ”the first conflict he experienced between the mother’s language and the father’s language ‫وتعرضه ملواقف وخبرات تكون له الرصيد املناسب من لغة األم مع تقدم مراحل نموه العمري‬ “exposing him to appropriate standpoints and experiences with /from(?) the mother tongue as he grew up”15 b) 2 instances introduced by the particle li-, and thus orthographically bivalent (as the assimilated definite article is suppressed in writing, unless marked with šadda): ‫يشارك األردن اليوم االثنني دول العالم االحتفال باليوم الدولي للغة األم‬ ”on Monday, Jordan joins the states all over the world in celebrating the International Mother Tongue Day! ‫[ للغة األم لدى صغار السن‬sic]‫ حيث يخصص الجانب األكر‬،‫األجزاء التي يوظفها املخ لدراسة اللغات‬ ”the parts which are activated by the brain in studying languages, where the largest (?) part is reserved for the mother tongue/language in the young ones” c) some instances of the (appositive) construction al-luġa al-ʾumm: ‫ ويضيع األطفال بني لغتني هشتني ال‬.‫عندنا تنتزع اللغة االم من املدارس والجامعات حرفا ومضمونا‬

15 A broader context would help the interpretation here.

224

Gunvor Mejdell

“the mother language/tongue is being completely withdrawn from our schools and universities, and children are lost between two weak languages/varieties“ ،‫واللغة األم في سورية وهي السامية الغربية‬ “[…] the mother language/tongue in Syria, which is West Semitic” .‫من خالل التعبير الشفوي مع االستئناس باللغة األم لتيسير الشروع في القراءة والكتابة باللغة العربية‬ “[…] through oral expression together with the familiarity of the mother language/tongue, in order to facilitate beginning to read and write in the Arabic language” ‫ ب "ل ي "عتمد على ال "لغة األم ب "شكل أس "اسي وب "عده "ا م "ن السه "ل على‬.‫ال ي "قوم على أس "اس ت "فوق ل "غة وت "عميمها‬ ‫الطفل أن يتعلم حتى أربع‬ “[…] does not rest on the superiority and spread of one language, but is fundamentally supported by the mother language/tongue, and thereafter it is easy for the child to learn even four [languages]” ‫كل قوم من االقوام له ان يزعم ان لغته هي اللغة األم لكل اللغات االخرى‬ “every nation will claim that its language is the mother language of all other languages” My final source on this issue is from the survey on language practices and language attitudes, conducted in Greater Cairo in 2013 by a research team based in Oslo16 (Kebede et al. 2013). About 2500 respondents were asked about the reference of the two terms, whether they covered the common spoken language (al-ʿāmmiyya), the standard language (al-fuṣḥā) or both: – meaning of luġat al-ʾumm: ʿāmmiyya (64%), fuṣḥā (25%), both (11%); – meaning of al-luġa al-ʾumm : fuṣḥā (73%), ʿāmmiyya (20%), both (7%). As can be seen, there is little agreement among respondents, as there has been among the sample voices discussed in this article, but still there is a tendency that corroborates the findings above. 4. Conclusion ’Mother tongue’ is a metalinguistic concept with clear ideological undertones. In a European context, it was used first to promote a vernacular variety as a written language competing with the ’high’ standard Latin, second, as a notion that strengthened the idea of a homogeneous and separate nation state, by creating affiliations between language and people/nation and political state16 Project The Ideology and Sociology of Language Change in the Arab World.

Luġat al-ʾumm and al-luġa al-ʾumm

225

hood (often erasing multilingual practices that existed within those states). In the Arabic context, ’mother tongue’ apparently was transferred to a cultural context and political movement that was Pan-Arab in orientation, and came to be used in support of an Arab nationalism which was super-regional and affiliated with the common ’high’ written language, al-ʿarabiyya (al-fuṣḥā). As certain writers recognize, luġat al-ʾumm has tended to be interpreted as luġat al-ʾumma (the language of the community/nation). This is certainly so with regard to the educational and cultural establishment, who continues to disregard – to erase from the dominant language ideology – the role of the Arabic vernaculars, the genuine ’mother tongues’ as defined by international dictionaries and institutions. The concept of ’mother tongue’ was adopted into the Arabic language in two variants: luġat al-ʾumm and al-luġa al-ʾumm. While there is to some extent overlapping between these terms in actual written usage, some make a distinction between the ʾiḍāfa construction as the ’mother’s language’ (mother tongue) and referring to the vernacular, and the appositive construction as ’mother language’ in the more ideological meaning, and referring to standard Arabic. This last usage, al-luġa al-ʾumm seems, from the data I have gathered, to be predominant. However, usage is not stable. And writers writing from a minority perspective, or in a multilingual context, use luġat al-ʾumm about these nonArabic varieties (Amazight, Kurdish, Suryani), when arguing for a role for these mother tongues in education and society. The expanding use in the last decades of vernacular Arabic (ʿāmmiyya, dārija) in written, even literary, practices notwithstanding – a role for these mother tongues in primary education is still largely an unspoken issue.

References ʿAbd al-Muḥsin, Yaḥyā Farġal. 2009. ”Al-ʿarabiyya ... al-luġa al-ʾumm wa- ʾaṯaruhā fī takwīn al-ṭifl.” Majallat Tawāṣul 10: 22-26. Davies, Eirlys and Bentahila, Abdelâli. 1989. ”On mother and other tongues: The Notion of Possession of a Language”. Lingua 8: 267-293. Eisele, John. 2003. ”Myth, Values, and Practice in the Representation of Arabic”. International Journal of the Sociology of Language 63: 43-49. Fasold, Ralph W. 1997. ”Motivations and Attitudes Influencing Vernacular Literacy: Four African Assessments.” in Tabouret-Keller, Andrée et al. (eds). 1997. Vernacular Literacy. A Re-Evaluation. (Oxford Studies in Anthropological Linguistics). Oxford: Clarendon Press.

226

Gunvor Mejdell

Gal, Susan. 1998. ”Multiplicity and contestation among linguistic ideologies.” Kathryn Woolard and Bambi Schieffelin (eds.):   Language Ideologies: Practice and Theory, 317-331. Oxford: Oxford University Press. Kebede et al. 2013. Language Change in Egypt: Social and Cultural Indicators Survey. A Tabulation Report. (FAFO-report 2013:39) Māhir, Muṣṭafā. 2009. ”Al-ʿalāqa bayn ‘al-tarjama’ wa-l-luġa al-ʾumm’”. ʾAwāṣir 2: 8-27. [Dirāsāt fī al-tarjama.] Cairo: Al-markaz al-qawmī li-l-tarjama. Mejdell, Gunvor. 2008. ”What is happening to luġatunā l-gamīla? Recent media representations and social practice in Egypt.” Journal of Arabic and Islamic Studies 8: 108–124. http://www.uib.no/jais/docs/vol8/v8_8_Mejdell_108-124.pdf. Suleiman, Yasir. 2013. ”Arabic Folk Linguistics between mother tongue and native language”, in: Jonathan Owens, (ed.): Oxford Handbook of Arabic Linguistics, 264-280. Oxford: Oxford University Press. Yildiz, Yasemin. 2012. Beyond the Mother Tongue. The Postmonolingual Condition. New York: Fordham University Press.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic1 Maria Persson, Lund University

Abstract This paper discusses ways in which the three main forms of the verb, and patterns of switching between them, are exploited for syntactic and pragmatic purposes in urban Gulf Arabic. Recent work on circumstantial qualifiers in Semitic, including Gulf Arabic, has brought into focus the function of switching between verb forms, here labelled ‘gram switching’, as a device for marking hierarchical relationships in clause combining (Isaksson, Kammensjö, and Persson 2009; Isaksson 2011; 2013; Persson 2014a; 2014b; 2015). The surveys have established that a switch, from one verb form to another or from a clause with a verbal gram to a verbless clause, is an alternative to conjunctions signalling subordination. This indicates that the use of varying verb forms in Gulf Arabic may be a considerably more dynamic part of narrative syntax than the simple description of each verb form may suggest. A prerequisite for the use of gram switching is that the language contains morphological forms with rather well-defined grammatical content. This paper further develops some key results from an extensive project on tense, mood and aspect marking in urban Gulf Arabic where, specifically, fairly well-defined aspectual values pertaining to the verb forms were identified. The paper presents a synthesis of previous findings and provides rich examples of how gram switching is used as a device for marking non-main clause linking, parallel to the use of more traditionally recognized modes of extra-verbal marking.

1. Aim and scope Analyses of a large database of spoken urban Gulf Arabic, on which the study is based, have shed light on the use of the basic verb forms. They have called attention to the low frequency with which verb-external marking is used for tense, mode and aspect in contemporary spoken urban Gulf Arabic. They have also shown that the suffix form and the prefix form of the urban Gulf Arabic verb have mainly aspectual values whereas the active participle is a nominal                                                                                                                         1 I wish to express my gratitude to Riksbankens Jubileumsfond (http://www.rj.se) for funding the research for this paper. I also wish to express my sincere thanks to Jan Retsö for valuable comments and thought provoking reflections on many of the views stated and examples discussed in this article.

228

Maria Persson

form expressing a state. (Persson 2008a; 2008c; Eades and Persson 2013). Once the usage of each individual verb form has been established, the next step is to study the way in which these verb forms interact in narrative text. Bybee and Dahl (1989) noted that grammatical morphemes, for which they coined the term ‘gram(s)’, play a vital role in signalling temporal and aspectual relations, but also grammatical and discourse structures. The term “morpheme” seems appropriate in view of the fact that grams are language specific2 and can each carry more than one meaning: they can be portmanteau and, for example, convey both modal and aspectual values. The use and meaning of a particular gram may also vary over time. The Gulf Arabic verb forms mentioned above and, at least to some extent, mood marking, are examples of such grams. In clause linking, the use of different verbal grams in the two clauses, a ‘gram switch’, may be one way of marking clausal relations. It has been shown that such gram switching; specifically a switch of verb forms from one clause to another but also a switch from a clause with a verbal gram to a verb-less clause, is used in this way, i.e. as a device for marking hierarchical relationships, in Semitic clause combining (Isaksson, Kammensjö, and Persson 2009; Isaksson 2011; 2013; Persson 2014a; 2014b; 2015). Such gram switching may complement, but often wholly replace, marking through other devices such as subordinating or coordinating conjunctions. A second, important observation has been the rich and varied use of chains of asyndetically juxtaposed verbs. This was highlighted in Isaksson, Kammensjö, and Persson (2009) and has received further attention in subsequent work (Kammensjö 2011; Persson 2013a). These asyndetic junctures comprise verb chains involved in grammaticalization processes where one of the verbs in a chain becomes a clausal modifier, as in raḥat tidrus, ‘(she) went (she) studies’ = ‘she started studying’ (auxiliary + main verb) (Firanescu 2003; 2008: 188); ambiguous junctures of verbs of the same form, as in aǧlis ašūf, ‘(I) sit (I) watch’ = ‘I sit and watch/keep watching/sit down to watch’ (Caubet 1996; Persson 2009: 265-269; 2013a), but also verb chains where a switch of verb form, or other types of gram switch, signal a digression from main-clause to non-main clause, as in ǧalas yaktub, ‘(he) sat (he) writes’ = ‘he sat (down) to write’ (Bybee, Perkins, and Pagliuca 1994; Isaksson, Kammensjö, and Persson 2009; Isaksson 2011; Persson 2012). All three examples, and the short definitions provided here, will be ex                                                                                                                         2 Bybee and Dahl (1989), however, postulate that (a few, major) “gram-types” are cross-linguistically valid.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

229

panded upon as I describe the theoretical background to this study. It is worth noting, however, that not only aǧlis ašūf but also the other two examples are ambiguous: raḥat tidrus, ‘(she) went (she) studies’ = ‘she went to study/she started studying’ (final clause or auxiliary + main verb) and ǧalas yaktub, ‘(he) sat (he) writes’ = ‘he sat (down) to write/ was writing/ sat writing’ (final clause or progressive or circumstance). In both of these examples, there is a shift of verb form from the suffix form to the prefix form. This kind of gram switching is the focus of this paper which will show how the above mentioned, rather well-defined, differences in aspectual values in the Gulf Arabic verb system, especially between the prefix form and the suffix form of the verb, are used to signal clausal and narrative hierarchy without recourse to conjunctions or discourse markers. More specifically, I will show that a switch between verb forms or mood marking is used in Gulf Arabic as an alternative to the use of extra-verbal markers of non-main clause linking such as complement clauses, conditionals, and final clauses. I also propose that this alternative way of marking non-main clause linking and narrative background is one reason why Arabic dialects are well-known for their common use of asyndetic clause combining. After a discussion of the theoretical background and the factors that motivated the study, I will present examples of the main functions of the simple verb forms, including a few mood markers that combine with the prefix form. I will, in other words, describe the main grams that operate in Gulf Arabic clause structure. This descriptive section has developed out of previous research and constitutes the basis for the discussion on the use of combinations of verb forms, and, particularly, the use of gram switching to signal sentence hierarchy and narrative foreground and background in clause combining and in discourse. 2. Previous studies Work on clause linking such as linking in the examples above has led to the insight that arriving at a clear definition of subordination is more complicated than the common use of the term may suggest (cf. Isaksson 2011; 2013). With this in mind, and following the works of Matthiessen and Thompson (1988) and Dixon (2009), clauses in this paper will simply be divided into main clauses and non-main clauses. The question of form and function in main-clause/non-main clause linking, and in particular the common use of asyndetic clause combining,

230

Maria Persson

constitutes a central but, as it seems, largely unresolved topic in Arabic and Semitic syntax. Clause combinations such as the last example above, ǧalas(a) yaktub(u), ‘(he) sat (he) writes’ = ‘he sat (down) to write/sat writing’, are well known from descriptions of Arabic grammar where they are mentioned among the so called ḥāl-clauses or circumstantial clauses (Abboud 1986). They are described together with longer circumstantial clauses such as ǧalasa wa-huwa yaktubu, ‘(he) sat and he writes’ = ‘he sat writing’. Undoubtedly, the same function of describing a circumstance may be fulfilled by both clauses. The question of how or why the shorter clause in ǧalasa yaktubu may also take on a final meaning ‘in order to write’, however, is largely left unanswered. Abboud discusses this clause combining at some length. He attempts to mediate between the presupposition that circumstantial clauses should encode simultaneity with the main clause event, and the fact that this is not the case in this type of clause combining; a type of clause combining that, nevertheless, is traditionally counted among the ḥāl-clauses. He suggests that this meaning is due to the aspect of the verb in the main clause, arguing that with ‘event-completion verbs’ the ḥālclause may refer to both the event itself and the completion of the event. In what is still called a circumstantial clause, the circumstance is said to be simultaneous, not with the action encoded in the verb, but with the event that constitutes the completion of this action. The final sense in “he sat down in order to write” is reinterpreted as ‘he sat down and, once seated, he was writing”. I agree with Abboud that the semantics of the verb in the main clause most likely plays an important role in deciding which verbs can combine in this type of main-clause/non-main clause linking. I believe, however, that this clause combining can and should be seen from a wider perspective than one of traditional theories on ḥāl-clauses. It is important to note here that ǧalasa yaktubu, in both its interpretations, comprises a main clause and a non-main clause. The only structural difference between the main clause and the non-main clause lies in the difference in verb forms. The semantic interpretation has to be derived from the wider context. Similarly, there are studies of Arabic verbs that are used as auxiliaries or that are in the process of changing from full verbs to grammaticalized markers of modal or aspectual value. In their bleached, grammaticalized uses, they have been shown to express aspectual values ranging from instantaneous to ingressive (Persson, 2013a). Introduced under headings such as “preverbs”, “prä-

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

231

modifizierende VP”, or “auxiliaries”, verbs such as qām ’stand up, get up’; rāḥ ‘go’ and ǧāʾ ‘come’ are among the most discussed, semantically bleached verbs in Arabic dialect (Woidich 2002; cf. Firanescu 2003; 2008). In such studies we find the other two examples mentioned above: a) raḥat tidrus, ‘(she) went (she) studies’, when used to mean ‘she started studying’, and b) aǧlis ašūf, ‘(I) sit (I) watch’, when used in the sense ‘I keep watching’. As illustrated above, these clause combinations also occur in descriptions of other parts of grammar. Such an alternative use of the combination is when they are intended to be interpreted as final clauses, i.e. ‘she went to study’ and ‘I sit down to watch’. Final non-main clauses of this structural type are, as mentioned above, traditionally treated in descriptions of circumstantial clauses (ḥāl-clauses). In such descriptions, as well, only the semantic interpretation of the combination in the particular context is considered. The function and identity of several of these clausal combinations as main/non-main clause linking are not in focus. Finally, constructions like these are at the centre of an on-going debate as to whether Arabic is a language with verb serialization or not (Versteegh 2009; Woidich 2002; Persson, 2013a). In other words, the typological classification of the types of linking exemplified above is, in many ways, an unresolved issue in Arabic linguistics. The clause combinations discussed here are all examples of asyndetic clause combining. That such asyndesis is common in Arabic has been noted by a number of researchers (Fischer 2002; Woidich 2002; Waltisberg 2009). Fischer (2002: 153, 161) suggests that asyndetic coordination of verbs of equal status is a characteristic of Arabic dialects that differentiates them from Classical Arabic. Nevertheless, two of the three examples mentioned above involve combinations of two different verb forms: the first verb is in the suffix form, the second one is in the prefix form. Only the last example is a combination of two verbs of the same form. Semantically, neither of the combinations can be said to constitute linking of clauses of “equal status”. They all involve a main clause and a clause that, at least in terms of its semantic content, is clearly not a main clause. Secondly, the extent of ambiguity is striking. All three clause combinations discussed in this section are found under more than one heading in grammatical descriptions of the language; most of these non-main clause types can encode more than one type of non-main clause and they do this without resorting to the use of subjunctions or conjunctions that clarify the relationship between the clauses. Such

232

Maria Persson

non-specific marking of non-main clauses through a switch of verb forms is, as has already been stated, the core topic of this paper. 3. Database The database used for this study comprises approximately 12 hours of fully transcribed recordings, and samples from a further 20 hours of not yet fully transcribed data. The complete database of urban Gulf Arabic represents 100 speakers from 25 different towns and villages in the Gulf area: Bahrain, United Arab Emirates, Kuwait, Oman, Qatar and Eastern Saudi Arabia. The interviews were conducted between 2006 and 2011 and mainly feature descriptions of customs, traditions and the difference between the old and the modern times, i.e. the main genre is informal narrative with interspersed dialogue. Previous research into this data has been presented in Eades and Persson 2013 and in Persson 2008a; 2008b; 2008c; 2009; 2013a and 2013b. 4. Verbal grams in Gulf Arabic The basic forms of the Gulf Arabic verb, the main verbal grams, will be labelled the prefix form, the suffix form, and the active participle. Certain mood markers will also be found to have a function as grams in the system. A gram switch is a switch from one of these grams to another or from a clause with a verbal gram to a verb-less clause. 4.1 Prefix form In terms of frequency of use in the data base, the prefix form of the verb is, by far, the most commonly used of the three verb forms. Its basic function is imperfective and, as such, it is used to signify tenseless states or activities. The time reference is relative, i.e. set by the discourse. This means that the prefix form is used for timeless statements of general truth; for present tense situations; for future events or activities as well as for imperfective events or activities in the past, i.e. the habitual past:3

                                                                                                                        3 Cf. Holes (2004, 219).

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

233

Imperfective, statements of general truth: (1) tištaġil work.PF5.3FS

fi-l-kahraba4 PREP-the-electricity

It works on electricity. Imperfective, present tense situation: (2) wald-i yadrus handasa son-my study.PF.3MS engineering My son studies engineering in France.

f-Faransa6 PREP-France

Imperfective, future tense situation: (3) baʿd ǝl-wilāda ahtamm after the-birth take_interest.PF.1CS

bi-l-bēbi7 PREP-the-baby

After the birth I will take care of the baby. Imperfective, past tense situation (habitual past): (4) iḥna ništaġǝl fi buyūt-na min PREP we work.PF.1CP PREP houses-our We used to work in our houses in olden days.

gabǝl8 before

The most interesting of these is, perhaps, the last example (4). Habitual aspect, as Östen Dahl (1985: 189) observed, plays an important role in the tense and aspect systems of many languages but is seldom marked as a separate category. Accordingly, in urban Gulf Arabic this use of the simple prefix form of the verb is the most common way of expressing a habitual action in the narrative past, once the time reference is established by an auxiliary or a time adverbial. (Persson 2008b, 2008c. Cf. Qafisheh 1977 on Gulf Arabic in general; and Johnstone 1961 on the Dōsiri dialect in Kuwait).

                                                                                                                        4 Elderly female, Fujairah, UAE. 5

”PF” in the examples stands for prefix [conjugation], ”SF” for ”suffix [conjugation].

6

Female teacher, Mubārak al-Kabīr, Kuwait

7

Young educated female, Fujairah, UAE.

8

Elderly female, Doha, Qatar.

234

Maria Persson

4.1.1 Mood markers 4.1.1.1 The Gulf Arabic b-prefix One mood marker that combines with the prefix form and that has received some attention in the literature is the Gulf Arabic b-prefix. As is the case with several other Arabic dialects, the prefix form of the Gulf Arabic verb may be combined with a b-prefix. Nonetheless, the function of the b-prefix in Gulf Arabic is radically different from the function of the b-prefix in Egyptian and Levantine dialects, for example (Retsö 2014; Holes 2004: 226-227; Brustad 2000: 241-253). The Gulf Arabic b-prefix has previously been assumed to signal future tense and/or intent (Johnstone 1967, 142ff; Brockett 1985: 21f.; Al-Maʿtūq 1986; Brustad 2000; Holes 2004: 247 n.29). A thorough analysis of the occurrences of this b-prefix in the present database (Persson 2008a) showed, however, that the b-prefix cannot be said to be a marker of future tense per se. Rather, the b-prefix is used for assumptions about a world that has not been experienced (futures, conditionals) or generalisations about past tendencies (habitual pasts): (5) ǝl-wāḥde b-itzawwaǧ bi-ṯalāṯīn ḫamse u ṯalāṯīn ʿādi and thirty normal the-one(F) IRR-marry.PF.3FS PREP-thirty five yaʿni šayʾ ṭabiʿi ʿādi ǝl-ḥīn9 normal the-time mean.PF.3MS thing natural For someone to marry at thirty, thirty-five is normal, you know, it’s a natural, normal thing now. In the case of most of the habitual pasts where the b- is used, the speaker has not him/herself experienced the past he/she is describing. The b- is used for the unlived, the non-evidential. In other words, the main function of this b-prefix in modern urban Gulf Arabic is as a marker of notionally irrealis categories.10 This, in turn, may (but need not) coincide with future temporal and/or intentional modal values. Neither future tense nor intentionality are signalled by the bprefix itself.

                                                                                                                        9 Young educated female, Doha, Qatar. 10 This does not imply that Gulf Arabic has a binary system for marking the realis/ irrealis distinction, but it does establish that urban Gulf Arabic has a marker that can be used to mark some events as ‘purely within the realm of thought, knowable only through imagination.’ (Mithun 1999: 173).

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

235

4.1.1.2 The particle rāḥ/ḥa The survey of future markers in urban Gulf Arabic in Persson (2008a) also established the use of another particle for future and, at least at times, irrealis contexts: the particle rāḥ/ḥa. In the following example, a young Emirati girl answers in this way to the question about how her life would have been if she had lived in the times of her grandmother: tikūn ḥayāt-i akīd miṯil ḥayāt-hum11 PTCL be.PF.3FS life-my certain like life-their For sure, my life would have been like their lives.

(6) rāḥ

The particle rāḥ/ḥa has received little attention in the literature on the Gulf Arabic dialects following Johnstone’s descriptions from 1967. In the Gulf Arabic data it is rarely used in non-future contexts and mainly found in Omani data (Persson 2008a: 35, 38). 4.1.1.3 The verb baġa/yabi Finally, a differentiation has been made between the b-prefix and the verb baġa/ yabi from which the prefix is, most often, thought to derive. The verb baġa/yabi has been noted to express both ‘proximate intention’ and ‘wanting’. In Gulf littoral dialects in general, Holes (2004: 247 n.29) has observed a distinction between ba-/bi- expressing intention (barūḥ ‘I’m going to go’) and yabbi + verb for volition (ʾabbi arūḥ ‘I’d like to go’). This verb has also been mentioned as indicator of future time in the adjacent Najdi dialect (Ingham 1994: 120f). In the present database, it is, in accordance with Holes’ observation, found to be used only for the volitive, i.e. to express a wish: (7) li-ann illi tabġi tikammil tigdar tikammil12 because REL want.PF.3FS finish.PF.3FS can.PF.3FS finish.PF.3FS For the one who wants to finish is able to finish.

                                                                                                                        11 Young educated female, Fujairah, UAE. 12 Young educated female, Doha, Qatar.

236

Maria Persson

(8) kānat

tabi

was.SF.3FS want.PF.3FS

šē13

tilabbis-na

aḥdaṯ

dress.PF.3FS-us

more_modern thing

She wanted to dress us in the latest fashion (lit. most modern thing). There were no instances in the database where the verb baġa/yabi was observed to signal future time-reference (Persson 2008a). 4.2 Suffix form The prefix form is, as mentioned, the verb form which is by far the one most exploited. The suffix form, on the other hand, is the one most limited in use and scope. It is mainly perfective, signalling completed action: (9) ǧibna ǝl-ʿayyāl u kabarna u ǝl-ḥīn grow.SF.1CP and the-time bring.SF.1CP the-children and naseyna14 forget.SF.1CP We got the children and grew (old) and now we have forgotten. The completed action can also have a future temporal setting, i.e. when the completion of the action is perceived as so strongly assured that for all practical purposes, it may be treated as if it were completed: ʿumr aš-šuyūḫ15 age the-sheikhs May God lengthen the lives of the sheikhs!

(10) ṭāl prolong.SF.3MS

Well known examples of this are oaths, invocations and expressions of a predestined future. 4.3 Active participle The active participle, formally, is a nominal. Hence, the clause in which it occurs as a predicate is verb-less and lacks tense/aspect morphology. The common generalization, that the active participle in dialects is a tense marker or at                                                                                                                         13 Middle aged, educated female, Shia, Manama, Bahrain. 14 Elderly female, Doha, Qatar. 15 Elderly female, Abu Dhabi, UAE.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

237

least a marker of perfect aspect, does not seem to hold for Gulf Arabic (Al-Tajir 1982: 111; Holes 1990: 22f., 191; Cuvalay-Haak 1997: 228, 240). Rather, the active participle only represents a state; a state that is caused by the activity encoded in the semantics of the underlying verb. As a nominal form expressing a state caused by an activity, the active participle can be used in a number of contexts with a wide array of meanings such as the non-progressive continuous, the perfect or resultative, and also future tense. The temporal or aspectual interpretation of a particular participle in a specific clause, however, is not inherent in the verb form, but derived from the interplay between the type of situation as determined from the context, the stative character of the active participle, and the lexical aspect value(s) of the verb stem.16 The interplay between these factors results in an inferred aspectual/temporal reading that is not signalled by the participial form in and by itself. (Eades and Persson 2013; Cf. Eksell 1985; Eisele 1990a, 1990b; Kinberg 1992). The stative character of the active participle means that it is often used to express the non-progressive continuous of stative verbs: Present/continuous (11) mumkin huwwe yaʿtamid ʿala aš-šaḫṣ wēn huwwa possible it depend.PF.3MS PREP the-person where he 17 ʿāyiš live.AP.MS It may depend on the individual; where he lives (viz. where he is in a state of living). A number of researchers have assigned a resultative, past perfect meaning to the active participle, except in the case of participles based on translocative verbs. The latter are also said to be able to express continuous and/or future activity. Eisele has pointed out, however, that these conclusions may be somewhat premature (Eisele 1999: 14-25, 127-147. Cf. also Eades and Persson 2013). It is true that active participles in context are often found to have a resultative                                                                                                                         16 Many predicates do not have a stable Aktionsart value. Rather, the interplay between grammatical form and context-induced situation type will, in context, bring one of several possible Aktionsart values to the fore. (Dahl 1985: 26f. Cf. Oldsjö 2001: 156, 170ff. and Croft 2009: 146ff.). 17 Young educated male, Barka, Oman.

238

Maria Persson

meaning. It is also the case that participles of translocatives are readily used in future contexts. Even so, as stated above, it is not the participial form itself that expresses all these aspectual and/or temporal values. When active participles in their context are found to have a resultative meaning this, too, is an inferred reading in a given instance of active participle use: Past perfect/resultative (12) fīh uġniya muʿayyina muš ḥāfiẓat-ha18 NEG memorize.AP.FS-it in.it song specific There is a certain song that I haven’t memorized. Example (12) may be paraphrased as “I am not in a state of having memorized it”. The addition of ‘a song’ to the verb ‘memorize’ makes the verb ‘memorize’ telic. This puts emphasis on the completion of the activity. Hence, the participle will signify the state of the agent (only) after the completion of the activity. This, then, is the reason behind its resultative, past perfect value in this context. Due to the stative aspectual meaning associated with the active participle, it also serves to express relevance to the speech moment, personal experience and other aspects of near connection between the state/activity expressed by the underlying verb and the reality of the speaker and the speech moment. This contributes to giving the participle a resultative/past perfect reading in context: (13) ana

baʿd yatīma umm-i mayyita min yōm ʿumr-i age-my I also orphan mother-my die.AP.FS from day ṯalāṯ ayyām19 three days I am an orphan too; my mother has been dead from when I was three days old.

The mother is dead. The fact that the mother died has an impact on the speaker’s life at the present moment. She is an orphan because of what happened; the past event is still a living reality for this woman. This is emphasised by the use of a stative form; by describing the death as a state rather than as an event.                                                                                                                         18 Young educated female, Doha, Qatar. 19 Elderly female, Doha, Qatar.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

(14) tšuft

ǧiddat-I mātat yimkin ʿind-ha grandmother-my die.SF.3FS maybe with-her

239

ṯamānīn sane20 eighty year

see.SF1.CS I saw my grandmother; she was maybe eighty years old when she died.21

In this second example the death of the grandmother is stated as a fact. The relevant information in the speech situation is her age when she died. The death in itself has relevance only for the time when it happened. Other temporal/aspectual contexts in which we find the active participle are pasts and futures: Past (15) sawwū-l-hum buḥūṯ qalū-l-hum liʾanna make.SF.3MP-PREP-them research say.SF.3MP-PREP-them because intu mʾāḫḏīn qarīb22 you.CP take.AP.MP close They examined them (and) said to them (it is) because you have taken relatives (in marriage). Future (16) fa-ana ṭāliʿ asās ykūn li … b-rūḥ-i23 so-I go_out.AP.MS base be.PF.3MS PREP-me … PREP-soul-my I’m leaving (the house) in order (for there) to be for … myself (a house). (17) inšalla umm-i msāfira l-ḥaǧǧ if_God_will mother-my travel.AP.FS the-pilgrimage 24 hiyya u abū-y she and father-my God willing my mother is going to go on the pilgrimage, she and my father (together).                                                                                                                         20 Middle aged, educated female, Shia, Manama, Bahrain. 21 A more idiomatic translation would be: ’I saw how/that my grandmother died at about 80 years of age. 22 Young educated female, Doha, Qatar. 23 Young educated male, Barka, Oman. 24 Young girl, Kuwait City, Kuwait.

240

Maria Persson

Futurate uses of the Arabic active participle are especially known to occur with locational verbs, such as the above. Eisele (1999: 139ff.) found, however, that participles derived from practically all verb stems, i.e. not only locationals, can be used in combination with future time reference in Cairene Arabic when the clause or context expresses intention and/or volition. Judging from the analyses of the data so far, this statement seems to be too strong for Gulf Arabic. Also in the Gulf Arabic database it is clear, however, that strong intention can license futurate readings even of participles formed from non-locational verb stems in hypothetical contexts where no intention whatsoever is involved: (18) … u

bukra

iḏa … baġēt

tištǝġǝl… … and tomorrow if … want.SF.2MS work.PF.2MS … ma qādir timsǝk ḥāge25 NEG be_able.AP.MS touch.PF.2MS thing … and tomorrow (i.e. in the future), if you … wanted to work … you wouldn’t/won’t be able to grab a thing.

This concludes the overview of the main verb forms, i.e. the main grams that may be used for processes of gram switching. I will now discuss examples of how the rather well-defined aspectual and modal values described here are used to encode clausal and narrative hierarchy. 5. Combinations of prefix and suffix forms. Gram switching to signal clausal and narrative hierarchy In the introduction to this paper, I referred to Fischer (2002: 153, 161), who proposes that the asyndetic coordination of verbs of equal status is a characteristic trait of Arabic dialects. According to Fischer’s observations, asyndetic verb coordination is the norm, the unmarked case in Arabic dialects whereas the use of coordinating particles or conjunctions produces marked cases. This observation, however, seems to convey only part of the full picture since asyndetic clause combining of verbs of unequal status is also a common phenomenon in the urban Gulf Arabic data.26 Frequently, both when the verbs are directly juxta                                                                                                                         25 Elderly female, Muscat, Oman. 26 As noted in the introductory section, many of the observations made in this paper are not only true of Gulf Arabic, but I will limit my statements here to the database on which the study is based.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

241

posed and when they are separated by other sentence parts, a shift of verb form is the sole indicator of clausal/narrative hierarchy. The rest of this paper will be devoted to a discussion of ways in which prefix and suffix forms are combined and, particularly, how a switch of verb form or a switch in mood marking is used in urban Gulf Arabic to encode non-main clause linking or narrative foreground/background. 5.1 General notes on the use of combinations of prefix and suffix forms A common combination is where a suffix verb is used at the onset of a description of habitual activities in the past. The most basic and best-known combination of this kind is the use of the auxiliary verb kān, ‘to be’, combined with the prefix form or the participle of the main verb (kān yafʿal/ kān fāʿil) to express the habitual or imagined past (19, 20), or to convey hypothetical statements about an alternative reality such as conditionals (21): (19) amma bi-n-nisba li-l-badu as.for PREP-the-relation PREP-the-Bedouin fa hum kānau yiḥṭabūn ǝl-ḥaṭab27 so they be.SF.3MP log.PF.3MP the-wood As for the Bedouins, they used to cut and collect wood. (20) kān yaba yirudd wi-yiʿīš be.SF.3MS want.PF.3MS return.PF.3MS and-live.PF.3MS fi-r-raml wa-ṣ-ṣaḥra28 PREP-the-sand and-the-desert He wanted (was wanting) to return and live in the sands and the desert. (21) kint agʿad wa uʿa aṣ-ṣubḥ wake_up.PF.1CS the-morning be.SF.1CS sit.PF.1CS and u ašūf ǝl-bēbe u aʿarḍ-ha li-š-šams expose.PF.1CS-her PREP-the-sun and see.PF1CS the-baby and assawi l-ha massaǧ uṣabbiḥ-ha29 make.PF.1CS PREP-her massage bathe.PF.1CS-her                                                                                                                         27 Middle aged male, Abu Dhabi and Dubai, UAE. 28 University student, male, Dubai, UAE. 29 Young, educated female, Abu Dhabi, UAE.

242

Maria Persson

I would be getting up in the morning and look at the baby and expose her to the sun, give her massage, bathe her.30 When used at the onset of a description of habitual activities in the past, the suffix verb, whether it is kān or another verb, is often perceived to have a temporal function in that it sets the time frame for the narration. This, however, is not really a temporal function but the same aspectual function that was observed with the bare form of the verb form. The scope of the suffix form (auxiliary) verb in 20 and in 21 covers more than the main verb: it functions as an aspect marker for the whole clause. Hence, the suffix verb, used in this way before a series of prefix verbs, signals that the succession of events described, seen as a whole, is completed: (22) aṣ-ṣeyd kān ǝl-arnab wa yiṣṭādūn ǝl-arānib hunt.PF.3MP the-rabbits the-hunt be.SF-3MS the-rabbit and yiṣṭādūn ǝl-... ṭēr ǝl-kurwān u ǝl-ḥabāri31 he-houbara_bustard hunt.PF.3MP the-... birds the-curlew and Hunting was rabbits, and they hunted (used to hunt) rabbits and they hunted (used to hunt)… birds: curlew(s) and (houbara) bustards. That the suffix verb does not set the time reference is even more evident when we see that it can also be placed after the prefix verb/verbs, as a kind of ending statement that confirms an already established perfective aspect for the preceding action/event series: (23) taʿrifīn ǝl-ummahāt yaʿni yištaġilūn šuġl kaṯīr know.PF.2FS the-mothers mean.PF.3MS work.PF.3CP work much kānu fi-l-bēt32 be.SF.CP PREP-the-house You know, the mothers used to do a lot of work in the house.

                                                                                                                        30 Said by the mother when asked what she would do if she were to stay at home with her baby instead of going to work. 31 Middle aged male, Abu Dhabi and Dubai, UAE 32 Young educated female, Doha, Qatar

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

243

Due to the rather specific aspectual differences between the verb forms, a combination of prefix forms and suffix forms in a sentence can be used to signal that the individual actions are seen from different perspectives: (24) ana

bi-nafsi PREP-self

ḥatta even

rās-i head-my

ma

mašṭ-eh brush.SF.1CS-it

NEG I 33 aḫāf astahi be_afraid.PF.1CS be_shy.PF.1CS I myself didn‘t even brush my hair; I was (used to be) afraid, I was (used to be) abashed.

Time in (24) is already set by the context. The whole discourse is in the past. The difference between the first predicate and the two later ones is that the combing of the hair is seen as comprising individual events, each instantaneous and finished, whereas the two second verbs describe a constant, ongoing feeling/ experience, albeit one that is also now a terminated experience. Similarly: (25) hay ʿayyāl-i ana rabbēt-hum these children-my I raise.SF.1CS-them kull-eh ṯiyāb-hum aġsal-eh34 all-it clothes-their wash.PF.1CS-it These my children, I have raised them washing all their clothes. The raising of the children is seen as a completed unity, a perfective whole. Washing clothes, on the other hand is seen as a repeated, iterative activity throughout this period. This habitual aspect is signalled by a switch to the prefix form. 5.2 Gram switching as a marker for clausal and narrative hierarchy The best-known type of clause linking, where a simple switch of verb forms is used to signal a hierarchical relationship between the clauses, are clauses expressing an activity or event that is simultaneous to the activity or event in the main clause, often labelled circumstantial:                                                                                                                         33 Elderly female, Doha, Qatar. 34 Elderly female, Doha, Qatar.

244

Maria Persson

(26) u and

ana

adrus

tzawwaǧt35

I

study.PF.1CS

marry.SF.1CS

I married while studying. This way of signalling non-parallel clause linking exploits the differences in aspectual values between the verb forms. When the bulk of the narration, the main-line of the story, is conveyed using one verb form, a sudden switch to another verb form will signal that this clause is not a main clause. In the following example, marking one of the clauses as a non-main clause only involves the insertion of an imperfective prefix verb into an otherwise perfective context: (27) waḥde sayyida timši fi š-šāriʿ saraqaw ḥaqībat-ha36 purse-her one lady walk.PF.3FS PREP the-street steal.SF.3CP A lady, who was/while she was walking in the street, they stole her handbag. By using the prefix form of the verb ‘walk’, timši, the narrator creates a nonmain clause that may be interpreted as temporal (circumstantial) “while she was walking” or relative “who was walking”. No other signal of subordination is needed and the narration continues with the suffix form to indicate that the whole context is in the completed past. Another clause type that is easily created by gram switching is the complement clause: ḫuftu tsawwu ḥādiṯ?37 NEG fear.SF.2CP make.PF.2CP accident Were you not afraid that you would have (lit. make) an accident?

(28) ma

The most common clause type where we find the use of gram switching in the database, however, is the conditional clause. Here is an example from the Emirati East coast: (29) ǝl-yōm the-day

hāḏa this

fulāna someone

ma NEG

arēt-ha see.SF.1CS-her

                                                                                                                        35 Young teacher, female, Kuwait. 36 Middle aged, educated female, Shia, Manama, Bahrain. 37 Young teacher, female, Buraimi, Oman.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

ana aba

arūḥ lē-ha go.PF.1CS PREP-her

245

ruḥt38 go.SF.1CS

I want.PF.1CS Today, this lady that I haven’t seen; if I want to see her, I go. There are no overt markers of condition in (29). A literal translation would be: “this day someone I haven’t seen her I want to go to her I went”. The switch between suffix and prefix forms is enough to signal a non-main clause linking. The context provides the conditional interpretation. A similar example is (30): (30) ana

muslim

ana b-adāfiʿ ʿan dīn-i (…) I Muslim I IRR-defend.PF.1CS PREP religion-my (…) waladti masīḥiyye bi-tdāfiʿīn ʿan ǝl-masīḥ39 be_born.SF.2FS Christian IRR-defend.PF.2FS PREP the-Christ I am a Muslim, I defend my religion (…) if you were born Christian you will/would defend Christ.

That a conditional interpretation of the gram switch is what the speaker intends is confirmed by his immediate restatement of the same idea using overt conditional markers: (31) lo kunti if be.SF.2FS lo kunti

muslima bi-tdāfiʿīn ʿan Muslim IRR-defend.PF.2FS PREP masīḥiyye ...40

ǝl-islām, the-islam,

if be.SF.2FS Christian… If you were a Muslim you would defend Islam, if you were a Christian … The informant in (30 and 31) is from Bahrain but has lived most of his life in Qatar. A Kuwaiti woman provides us with another example of how the insertion of a suffix verb in an otherwise prefix form verb context will create a conditional:

                                                                                                                        38 Elderly female, Fujairah, UAE. 39 Elderly male with home and family in Bahrain but working most of his life in Doha, Qatar. 40 Elderly male with home and family in Bahrain but working most of his life in Doha, Qatar.

246

Maria Persson

(32) Allah (…) ma God (…)

NEG

ha-l-makān

raḥ yiḫalli-ha

tḫalliṣ.

FUT let.PF.3MS-it

end.PF.3FS.

ḫallaṣ

biʾr well

ṯān yiṭlaʿ41 second appear.PF.3MS

This-the-place end.SF.3MS God (…) will not let it run out. (If/when) this place runs out a second well appears. Finally, another example from the UAE, this time from the West coast: (33) kān… be.SF.3MS… ǝl-ḥarīm the-women yitġaššūn43 veil.PF.3CP

lamma yiṭlaʿūn aš-šuyūḫ (...42) fi-t-televizion when appear.PF.3CP the-sheikhs (…) PREP-the-television ʿala ṭūl yitəәġaššūn, šēḫ, ṭalaʿ šēḫ, immediately veil.PF.3CP sheikh appear.SF.3MS sheikh

It used to be when sheikhs would appear… on TV, the women would immediately veil; a sheikh, if a sheikh appeared, they would veil. The first verb, kān, indicates the perfective aspect for the whole context. After the conjunction lamma, ‘when’, the following verbs are in the prefix form indicating ongoing, habitual past. This is followed by a suffix verb, ṭalaʿ, ‘he appeared’. No other marking is needed to indicate that this is a temporal-conditional: “if/when a sheikh would appear, they would veil.” The order of the two verb forms can also be the opposite: (34) hāḏa l-ḥukūme yigūlu l-i ǝl-yōm ṭilʿi leave.IMP.FS this the-government say.PF.3CP PREP-me the-day ṭalʿat, amr ǝl-ḥukūme hāḏa44 leave.SF.1CS, order the-government this This is the government, if they say to me today “leave!”, I leave; this is an order from the government!                                                                                                                         41 Middle aged female, Kuwait. 42 The informant quotes names of sheikhs. 43 Middle aged male, Abu Dhabi and Dubai, UAE. 44 Elderly female, Abu Dhabi, UAE.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

247

In narratives, this shift between verb forms also serves to differentiate between actions that are foregrounded and states or activities that are in the background. The gram switch is, then, a tool employed to separate between events that move the story forward and those that create the background scenery. Example (35) features a woman talking about the preparations for and the proceedings of a wedding and she continuously uses the suffix form to mark a progression in the story: (35) әyibūn

ǝl-maʿres ḫalāṣ bring.PF.3MP he-groom finished istawa hagg ʿars-hum istawa. concerning wedding-their happen.SF.3MS. happen.SF.3MS bēt ǝl-maʿres әhnāk yidiggūn ʿars House the-groom there play_drums.PF.3MP wedding u-huna bēt ǝl-ʿars yidiggūn45 and-here house the-wedding/bride play_drums.PF.3MP They (would) bring the groom. That’s it. Their wedding has come to happen, it has happened. (In) the groom’s house they (would) party (for a) wedding and here (in) the bride’s house they (would) party.

During the whole of this (rather long) description, the habitual past with prefix forms of the verb describes the traditions and customs at each stage of the preparation for the wedding. These prefix forms alternate with suffix forms that are inserted to signal the end of each stage and the progress to the next. In the example, we come to the end of the preparation process. When all the stages in the preparation are complete, the day of the wedding has arrived. This progression along the time line is marked by a suffix form, istawa. Then, the descriptive narrative continues with prefix forms which now indicate what would usually happen on the actual wedding day(s). The background element may also be a comment outside of the actual narrative such as an explanation added by the narrator:

                                                                                                                        45 Elderly female, Doha, Qatar.

248

Maria Persson

(36) ǧāt

šarika

ana

come.SF.3FS

company I

ǧāt

šarika …46 company

ahl-i

yiḫabbirū-ni

family-my inform.PF.3MP-me

come.SF.3FS A business company came, my family told me [this], a company came … In the story of what happened in the past, the narrator inserts a reference, telling the listener the source of his information. This extra piece of information is set apart from the main story with the help of a gram switch. Thus, the gram switch signals a temporary digression from the main line of the narration. These were isolated instances. The next example will serve to illustrate how this gram switching tool is used in context. Example (37) is a somewhat longer passage narrated shortly before (33) by the same informant: (37) fa kānun eee ǧaddati Allah yirḥam-ha tḫabbir-ni yaʿni tgūl: niḥna nrūḥ; yiḥuṭṭūn-hum sēnema; iḏa ṭallaʿ ḥarīm hum šallau әl-ġišwa, yitġaššūn; lamma yiṭlaʿ raǧul fi-s-sēnema fa-hum yiḍʿūn әl-ġišwa; yiḫūz әr-raǧul šallau әl-ġišwa; fa hum yistaḥūn47 The narration starts with the suffix form of, kān, in this case kānun, indicating the past perfective for the whole discourse. The following verbs, which constitute reported speech of someone deceased, are in the prefix imperfective form, indicating the habitual past: ǧaddati (…)

tḫabbir-ni yaʿni tgūl my grandmother (…) inform.PF.3FS-me mean.PF.3MS say.PF.3FS My grandmother, (..),used to tell me, you know, she’d say: The narration referred to is also in the habitual past: niḥna nrūḥ we go.PF.1CP We would go.                                                                                                                         46 Middle aged male, Abu Dhabi and Dubai, UAE. 47 Middle aged male, Abu Dhabi and Dubai, UAE.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

249

Then the informant adds a comment in the same manner: yiḥuṭṭūn-hum sēnema put.PF.3MP-them cinema They would put up a cinema for them. After this comes an explicitly marked conditional. The conditional is a context where the use of the suffix form is common: iḏa ṭallaʿ

ḥarīm hum šallau they take.off.SF.3CP if appear.SF.3MS women If women showed (on the screen) they unveiled.

әl-ġišwa the-veil

After that, some explanatory comments are added. Again, the habitual past is indicated by a reversal from suffix form to prefix form. No other marking: yitġaššūn cover.PF.3CP They used to veil. Then back to the narration with habitual pasts: Lamma yiṭlaʿ raǧul fi-s-sēnema fa-hum yiḍʿūn әl-ġišwa when appear.PF.3MS man in the cinema so-they don.PF.3CP the-veil When a man would appear on the cinema they would don the veil. A simple switch of verb forms in the following clause indicates the apodosis of a new conditional, one that is not introduced by a conditional marker: yiḫūz

әr-raǧul šallau әl-ġišwa the-veil disappear.PF.3MS the-man take.off.SF.3CP [If] the man would disappear they would unveil Finally, the narration is concluded with a statement in the habitual past: fa hum so they

yistaḥūn be_shy.PF.3CP

250

Maria Persson

because they would be abashed. Without this context, the suffix verb, šallau, would be interpreted as ‘they took off’. Here, when inserted in a narrative where the main story is told using verbs in the prefix form, it receives a conditional interpretation. The gram switch, the change from prefix form to suffix form, is the only marker of this conditional, the only marker of this non-parallel clause linking. Not only switches between verb forms but also changes in mood marking can be used in this way: (38) bi-n-nisba ḥagg PREP-the-relation POSS b-azawwiǧ-ha IRR-marry_off.PF.1CS-her b-azawwiǧ-ha48

bint-i daughter-my ayyi ḥadd any one

twaṣṣil iṯnaʿaš reach.PF.3FS twelve yīǧī-ha come.PF.3MS-her

IRR-marry_off.PF.1CS-her

As for my daughter, when she reaches 12 (years) I’ll marry her off, to anyone who comes to her (to ask her hand) I’ll marry her off. The temporal-conditional clause and the relative clause are marked by a lack of the irrealis marker. The daughter is expected to become twelve years old and ‘when or as she does indeed reach twelve’, the mother jokingly says, ‘I will marry her off’, to ‘anyone’ who is indeed expected to come. If all verbs had had the b-prefix, the meaning would have been more like ‘she will become twelve and I will marry her off. Anyone will come to her and I will marry her off’, i.e. perfectly parallel clauses; coordination, not subordination. This, then, is another type of gram switch, not between verb forms but a switch in mood marking. A switch between verbs with and without the irrealis marker can also be used to encode a consequence: a clause containing the logical result of the activity in the main clause: (39) b-adawwir ayy šuġl b-aštaġil-ah IRR-seek.PF.1CS any work IRR-work.PF.1CS-it u anta tartāḥ u umm-i tartāḥ49 and mother-my rest.PF.3FS and you rest.PF.2MS                                                                                                                         48 Young educated female, Doha, Qatar. 49 Elderly female, Muscat, Oman.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

251

I will seek any job and work in it and (then) you will rest and my mother will rest. (40) yimkin nanṭīr u nidʿam ǧabal ba-nmūt50 hit.PF.1CP mountain IRR-die.PF.1CP possible fly.PF.1CP and Maybe we fly and hit a mountain, (then) we will/would die. A gram switch with similar function is the imperative plus a prefix form: (41) lamma tūṣal ǝl-ʿēn ḫabbir fulān u fulān when arrive.PF.2MS al-Ain inform.IMP.MS someone and someone 51 yiǧū-na asās yiwaddū-na ǝl-ʿēn come.PF.3MP-us base bring.PF.3MP-us al-Ain When you arrive in al-Ain tell so-and-so that they should/may come to us to take us to al-Ain. The non-main clause following the imperative may be interpreted either as a complement clause, ‘inform him that they should come’; or as a consequence ‘inform him so that they may come’. A similar example from Bahrain: (42) gūlu li umm-kum tilabbis-kum ṭawīl52 say.IMP.CP PREP mother-your dress.PF.3FS-you long Tell your mother to dress you in long(-sleeved) clothes! And, finally, an example of switching between verb forms that contains the active participle: әṣ-ṣubəәḥ PREP the-morning daily bi-ṯiyāb-hum yəәġasslūn53 PREP-clothes-their wash.PF.3CP

(43) yawmīan min

tlāgēn find.PF.2FS

rāyḥīn go.AP.CP

                                                                                                                        50 Young teacher, female, Buraimi, Oman. 51 Middle aged male, Abu Dhabi and Dubai, UAE. 52 Middle aged, educated female, Shia, Manama, Bahrain. 53 Middle aged, educated female, Shia, Manama, Bahrain.

252

Maria Persson

Daily, from the morning, you find they’re gone with their clothes to wash/ washing. In (43) the context is in the habitual past. Every day, you would find them having gone to the well. Only a change from prefix form to the active participle, from tlāgēn to rāyḥīn, indicates the complement clause: that they are “in a state of having gone”. A new change of verb form, back to the prefix form yəәġasslūn indicates another non-main clause: either a final clause encoding the reason for their going: ‘in order to wash’; or possibly the circumstances in which you find them, namely ‘being in the process of washing’. No conjunctions or other external markers are needed to signal the hierarchy between the clauses; only a simple switch between verb forms. As we have seen also in other examples, the exact interpretation or function of the non-main clause is left to be interpreted from the context. 5.3 Discussion We have seen how suffix forms are used at the beginning of a narration to set the aspectual value for the whole context; suffix forms signal a progression on the time line in a narration that is mainly related in the habitual past with prefix forms; suffix forms are also introduced in a context of prefix verbs to signal temporal/conditional clauses. Prefix forms are inserted after suffix forms or participles to signal relative clauses, final clauses or conditional clauses. Similarly, prefix forms without mood marking (b-prefix) may alternate with prefix forms with mood marking to signal temporal/conditional clauses and relative clauses. Finally, active participle forms may also be used to signal a digression from the main clause, i.e. non-main clause linking. These all constitute examples of how gram switching is used in the modern urban Gulf Arabic database to signal clausal and narrative hierarchy. Gram switching is used as an alternative to the overt marking of non-main clause linking. The main difference being that the type of non-main clause is not explicitly stated but inferred from its context. As some of the examples have shown, conditional clauses with gram switching as the only marker of nonmain clause linking are sometimes found in the direct vicinity of clauses with explicit conditional markers. If clauses in which gram switching alone signals non-parallel clause linking were always found to follow clauses with overt markers of the same linking, this lack of an overt marker could simply have been re-

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

253

ferred to principles of economy. A marker used shortly before need not be repeated. The order, however, may just as well be the opposite, i.e. with gram switching alone being used before a clause where the clause linking is signalled by a combination of extra-verbal marking and a gram switch. Moreover, there are instances where neither gram switching, nor conjunction/subjunction is used. These are instances where the interpretation of one of the clauses as a non-main clause seems to totally rely on context. They are instances where the speaker does not seem to see a need even to clarify whether a clause should be interpreted as a main clause or a non-main clause; and if the latter, what type of non-main clause. Consider the following: (44) u-yōm ṯāni yәfrәšūn yigʿadūn yākәlūn-әh and-day second set.table.PF.3CP sit.PF.3CP eat.PF.3CP-it And the next day they set the table they sit they eat it. This series of three verbs may be understood sequentially: ‘they set the table,54 then sit down, then eat’. Since the verb gaʿad serves as a progressive marker in Gulf Arabic, the last two verbs could, however, also be interpreted as a progressive marker (yigʿadūn) plus a main verb, (yākәlūn): ‘they set the table and are eating’. The sequence between the first two verbs is inferred from common sense: one sets the table and then eats. Hence, it is not explicitly expressed. A third alternative interpretation of this sentence is to read the last verb as a circumstantial clause with either a final interpretation: ‘they set the table and/then sit down in order to eat’ or (manner) adverbial interpretation: ‘they set the table and/then sit eating’. In an actual speech situation, factors such as context, intonation, gestures etc. will often indicate the intended meaning of the sentence (Haspelmath 1995: 30; cf. Persson 2009: 258f). Still, the amount of ambiguity that the speaker seems comfortable with is striking. If no marking is needed, what may the function of the overt marking be? Or, if it is not needed, then why is it used? Perhaps there is no answer to this question. Perhaps it is a matter of style and speakers’ preferences. Previous studies on asyndetic clause combining and circumstantial clause combining in Arabic have discussed examples such as (44) and other areas where the language allows for rather far-reaching ambiguity in how to interpret the type of linking between main clauses and non-main clauses. As pointed out in Isaksson, Kammensjö, and Persson 2009, one of the hall                                                                                                                         54 More specifically in this context: spread out the dinner mat on the floor/ground.

254

Maria Persson

marks of what have been called circumstantial clauses, for example, is that they modify their main clauses in a way similar to adverbial, final, causal, conditional etc. clauses without explicitly marking the exact character of this relationship. The use of gram switching to encode various types of non-main clause linking seems to be an overarching structure and part of a general scale for marking non-main clause linking. Overall in the language, semantically non-specific means of marking non-main clause linking are used where context is enough to convey the intended meaning or where the speaker does not see a need for more explicit marking. With this knowledge, there is no need to force clauses, such as raḥat tidrus and ǧalas(a) yaktub(u), with final meaning into the proposed category of ḥālclauses, i.e. there is no need to postulate a ḥāl muqaddar with a “circumstance that holds at the completion of an action/event”. Ǧalas(a) yaktub(u) is a main clause/non-main clause combination signalled by a gram switch. The non-main clause has either an adverbial or a final value; ‘he sat writing’ or ‘he sat down to write’. The intended interpretation of the main verb and, hence, the semantic value of the non-main clause, is inferred from the context.55 6. Summary This paper has been divided into two sections; one descriptive section that builds on previous research and presents the verbal grams under study, and one analytic section where the use, in spoken urban Gulf Arabic, of gram switching to mark clause hierarchy has been surveyed and discussed. The first section confirmed that the suffix form and the prefix form of the urban Gulf Arabic verb have mainly aspectual values. The prefix form of the Gulf Arabic verb is imperfective, whereas the suffix form is perfective. The active participle is tenseless and aspectless and describes a state. The urban Gulf Arabic b-prefix, finally, was identified as a marker of irrealis modality. The findings of the second section of the paper suggest that the fairly welldefined aspectual values associated with Gulf Arabic verb forms, especially the prefix form and the suffix form, are employed in various ways in clause combining to signal clause hierarchy and discourse structure. This is done without recourse to conjunctions or discourse markers. Simple switches between grams                                                                                                                         55 In Gulf Arabic the verb ǧalas(a) is also used as an auxiliary and, hence, a third interpretation is possible: ‘he was writing’.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

255

(verb forms) are used to signal temporal/conditional clauses, relative clauses, final clauses and other non-main clause linking as well as narrative background/foreground. This kind of gram switching is sometimes used in combination with extra-verbal marking of main clause/non-main clause linking. Sometimes, it is used alone as an alternative to extra-verbal marking. This suggests that gram switching is the main marker for main clause/non-main clause hierarchy and that extra-verbal markers are an added convenience for clarification and specificity. More study on non-main clauses with extra-verbal marking is needed before definite conclusions on this matter can be drawn. Studies so far have concentrated on discovering the marking in clauses that lack extra-verbal markers. The use of gram switching only, i.e. when there are no conjunctions or subjunctions to clarify the type of non-main clause, was compared to earlier findings of how what have been called circumstantial clauses function as semantically non-specified types of adverbial or other non-main clauses. From the findings presented in this paper and from similar studies (Isaksson, Kammensjö, and Persson 2009; Isaksson 2011; Persson 2013a; 2014a; 2014b; 2015), a pattern is emerging. The observations work together to reveal that gram switching is an important way to signal a main clause/non-main clause juncture, the function of which is inferred from context. One important conclusion to be drawn from this is that there is no need to postulate a ḥāl muqaddar. Moreover, the high frequency of asyndetically juxtaposed verbs in Gulf Arabic, i.e. the apparent lack of conjunctions and subjunctions in many combined clauses, may partly be explained by this presence of another marker of clausal hierarchy.

References Abboud, Peter. 1986. “The ḥāl construction and the main verb in the sentence.” The Fergusonian impact. In honor of Charles A. Ferguson on the occasion of his 65th birthday. Vol. 1, From phonology to society, edited by Joshua A. Fishman, Andrée TabouretKeller, Michael Clyne, Bhariraju Krishnamurti and Mohammed Abdulaziz, 191196. Berlin: Mouton de Gruyter. Al-Maʿtūq, Šarīfa. 1986. Lahǧat al-ʿaǧmān fi l-Kuwait. Dirāsa luġawiyya. Doha: Markaz atturāṯ aš-šaʿbi li-duwal al-ḫalīǧ al-ʿarabiyya. Al-Tajir, Mahdi Abdalla. 1982. Language and linguistic origins in Baḥrain. The Baḥārnah dialect of Arabic, London: Kegan Paul International. Brockett, Adrian A. 1985. The spoken Arabic of Khābūra on the Bāṭina of Oman, Journal of Semitic studies. Manchester: University of Manchester.

256

Maria Persson

Brustad, Kristen Elizabeth. 2000. The syntax of spoken Arabic. A comparative study of Moroccan, Egyptian, Syrian, and Kuwaiti dialects. Washington: Georgetown University Press. Bybee, Joan L., and Östen Dahl. 1989. “The creation of tense and aspect systems in the languages of the world.” Studies in language 13/1: 51-103. Bybee, Joan L., Revere D. Perkins, and William Pagliuca. 1994. The evolution of grammar: tense, aspect, and modality in the languages of the world. Chicago: The University of Chicago Press. Caubet, Dominique. 1996. “Gālǝs kayxdǝm, xāyǝḍ kayxdǝm: Approche sociolinguistique de l’expression de la concomitance en arabe marocain.” Estudios de dialectología norteafricana y andalusí, edited by Jordi Aguadé, Federico Corriente, María José Cervera and Ignacio Ferrando, 87-100. Zaragoza: Universidad. Croft, William. 2009. “Aspectual and causal structure in event representations.” Routes to language. Studies in honor of Melissa Bowerman, edited by Virginia C. MuellerGathercole, 139-168. New York: Psychology Press. Cuvalay-Haak, Martine. 1997. The verb in literary and colloquial Arabic, Functional grammar series 19, Berlin: Mouton de Gruyter. Dahl, Östen. 1985. Tense and aspect systems. Oxford: Blackwell. Dixon, Robert M. W. 2009. “The Semantics of Clause Linking in Typological Perspective.” The Semantics of Clause Linking. A cross-linguistic typology, edited by Robert M. W. Dixon and Alexandra Y. Aikhenvald, 1-55. Oxford: Oxford University Press. Eades, Domenyk, and Maria Persson. 2013. “Aktionsart, word form and context: On the use of the active participle in Gulf Arabic dialects.” Journal of Semitic Studies 57/2: 343-367. Eisele, John C. 1990a. “Time Reference, Tense and Formal Aspect in Cairene Arabic.” Perspectives on Arabic Linguistics I. Papers from the First Annual Symposium on Arabic Linguistics, edited by Mushira Eid, 173-212. Amsterdam: John Benjamins. — 1990b. “Aspectual Classification of Verbs in Cairene Arabic.” Perspectives on Arabic linguistics II. Papers from the Second Annual Symposium on Arabic Linguistics, edited by M. Eid and J. McCarthy. Amsterdam: John Benjamins, 192-233. — 1999. Arabic verbs in time: tense and aspect in Cairene Arabic. Wiesbaden: Harrassowitz. Eksell, Kerstin. 1985. “On the function of the verbal active participle in northern Arabian narrative texts.” Acta Orientalia 46: 7-22. Firanescu, Daniela Rodica. 2003. “Le modalisateur aspectuel-temporel qām dans le parler syrien.” Proceedings of the 5th Conference of AIDA, held in Cádiz September 2002, edited by Ignacio Ferrando and Juan J. Sánchez Sandoval, 481-492. Cádiz: Service Servicio de Publicationes, Universidad de Cádiz. — 2008. “The moving sands of the modals raaH and ija in Syrian Arabic.” In Between the Atlantic and Indian Oceans. Studies on contemporary Arabic dialects. Proceedings of the 7th AIDA Conference, held in Vienna from 5-9 September 2006, edited by Stephan Procházka and Veronika Ritt-Benmimoun, 185-194. Wien: LIT Verlag.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

257

Fischer, Wolfdietrich. 2002. „Unterordnende und nebenordnende Verbalkomposita in den neuarabischen Dialekten und im Schriftarabischen.“ "Sprich doch mit deinen Knechten aramäisch, wir verstehen es!": 60 Beiträge zur Semitistik: Festschrift für Otto Jastrow zum 60. Geburtstag, edited by Werner Arnold and Hartmut Bobzin, 147-163. Wiesbaden: Harrassowitz. Grotzfeld, Heinz. 1965. Syrisch-arabische Grammatik: (Dialekt von Damaskus). Wiesbaden: Harrassowitz. Haspelmath, Martin. 1995. “The converb as a cross-linguistically valid category.” Converbs in cross-linguistic perspective: structure and meaning of adverbial verb forms – adverbial participles, gerunds, edited by Martin Haspelmath and Ekkehard König, 155. Berlin: Mouton de Gruyter. Holes, Clive. 1990. Gulf Arabic. London: Routledge. — 2004. Modern Arabic: structures, functions, and varieties. Washington, D.C.: Georgetown University Press. Ingham, Bruce. 1994. Najdi Arabic: central Arabian. Amsterdam and Philadelphia: John Benjamins. Isaksson, Bo. 2011. “The textlinguistics of the Suffering Servant: Subordinate structures in Isaiah 52,13–53,12”. En pāsē grammatikē kai sophiā. Saggi di linguistica ebraica in onore di Alviero Nicacci, ofm, edited by G. G. a. Massimo Pazzini, 173-212. Jerusalem and Milano: Franciscan Printing Press and Edizioni Terra Santa. — 2013. “Subordination: Biblical Hebrew.” Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan, vol. 3, 657-664. Leiden: Brill. Isaksson, Bo, Heléne Kammensjö, and Maria Persson. 2009. Circumstantial qualifiers in Semitic: the case of Arabic and Hebrew. Wiesbaden: Harrassowitz. Johnstone, Thomas M. 1961. “Some characteristics of the Dōsiri dialect of Arabic as spoken in Kuwait.” Bulletin of the School of Oriental and African Studies XXIV, Part 2: 249-297. — 1967. Eastern Arabian dialect studies, London oriental series 17. London: Oxford University Press. Kammensjö, Heléne. 2011. “Asyndetical clause combining and verbal gram switching in rural Egyptian Arabic.” Paper presented at Workshop on Circumstantial Clause Combining in Semitic Languages, 17-18 March 2011, Uppsala University. Kinberg, Naphtali. 1992. “Semi-imperfectives and imperfectives: A case study of aspect and tense in Arabic participal clauses.” Lingua 86/4: 301-330. Matthiessen, Christian, and Sandra A. Thompson. 1988. “The structure of discourse and ‘subordination’.” Clause combining in grammar and discourse, edited by John Haiman and Sandra A. Thompson, 275-329. Amsterdam: John Benjamins. Mithun, Marianne. 1999. The languages of Native North America. Cambridge: Cambridge University Press. Oldsjö, Fredrik. 2001. Tense and aspect in Caesar's narrative, Studia Latina Upsaliensia 26. Uppsala: Acta Universitatis Upsaliensis.

258

Maria Persson

Persson, Maria. 2008a. “The Role of the b-prefix in Gulf Arabic dialects as a marker of future, intent and/or irrealis.” Journal of Arabic and Islamic Studies 8/4: 26-52. — 2008b. “Progressiv och habituell aspekt i gulfarabiska.” (Progressive and habitual aspect in Gulf Arabic). Paper presented at Nordiska semitistsymposiet, 13-16/8 2008. Kivik, Sweden. — 2008c. “An interesting typological compromise. Report from a corpus-based study of modal and aspectual markers in Gulf Arabic dialects.” Paper presented at AIDA 8, 28-31/8 2008. Essex University, Colchester. — 2009. “Circumstantial qualifiers in Gulf Arabic dialects.” Circumstantial qualifiers in Semitic: the case of Arabic and Hebrew, Bo Isaksson, Heléne Kammensjö and Maria Persson, 206-289. Wiesbaden: Harrassowitz. — 2011. “Semantic and syntactic considerations in asyndetic circumstantial clause combining in Damascene Arabic.” Paper presented at Workshop on Circumstantial Clause Combining in Semitic Language, 17-18 March 2011, Uppsala University. — 2012. “Circumstantial clause combining and gram switch in Syrian Arabic.” Paper presented at International Symposium on Clause Linking in Semitic Languages 5-7/8 2012, Kivik, Sweden. — 2013a. Asyndetic clause combining in Gulf Arabic dialects. Auxiliary, adverbial and discourse functions. Zeitschrift für Arabische Linguistik 57: 5-39. — 2013b. “Verb form switch as a general marker of hypotax in Arabic dialects”. Paper presented at The 10th Conference of AIDA, 10-13/11 2013, Doha, Qatar. — 2014a. “Verb form switch as a marker of discourse hierarchy in Semitic: a case study on Syrian Arabic.” From Tur Abdin to Hadramawt. Semitic Studies. Festschrift in Honour of Bo Isaksson on the occasion of his retirement, edited by Tal Davidovich, Ablahad Lahdo and Torkel Lindquist, 117-128. Wiesbaden; Harrassowitz. — 2014b. “Non-main clause linking and verb form switch in Syrian Arabic. Is there a circumstantial clause?” Strategies of Clause Linking in Semitic Languages. Proceedings of the International Symposium on Clause Linking in Semitic Languages, Kivik, Sweden, 5–7 August 2012 (Abhandlungen für die Kunde des Morgenlandes 93), edited by Bo Isaksson and Maria Persson, 27-50. Wiesbaden: Harrassowitz. — 2015. “Non-mail clause combining in Damascene Arabic: A scale of markedness.” Clause Combining in Semitic. The Circumstantial Clause and Beyond. (Abhandlungen für die Kunde des Morgenlandes 96), edited by Bo Isaksson and Maria Persson, 55124. Wiesbaden: Harrassowitz. Qafisheh, Hamdi A. 1977. A short reference grammar of Gulf Arabic. Tucson, Arizona: The University of Arizona Press. Retsö, Jan. 2013. “The b-imperfect once again: typological and comparative perspectives.” Proceedings of the Oslo-Austin Workshop in Semitic Linguistics (Abhandlungen für die Kunde des Morgenlandes 88), edited by Lutz Edzard and John Huehnergard, 64-72. Wiesbaden: Harrassowitz. Waltisberg, Michael. 2009. Satzkomplex und Funktion. Syndese und Asyndese im Althocharabischen. Wiesbaden: Harrassowitz.

Verb form switch as a marker of clausal hierarchies in urban Gulf Arabic

259

Versteegh, Kees. 2009. “Serial Verbs.” In Kees Versteegh (ed.). Encyclopedia of Arabic language and linguistics, vol. IV, 195-199. Leiden: Brill. Woidich, Manfred. 2002. “Verbalphrasen mit asyndetischem Perfekt im ÄgyptischArabischen.“ Estudios de Dialectologia Norteafricana y Andalusi 6: 121-192.

Agglutinated verb forms in the Northern province of Yemen Ori Shachmon, The Hebrew University of Jerusalem

The Arabic dialects of the northern province of Yemen show extreme archaisms alongside a host of innovative features. Among the latter one can point to the development of agglutinated verbal forms in the perfect paradigm. This paper aims to illuminate these unique verbal structures, namely the addition and subsequent agglutination of the independent pronoun ant/int to the basic 2.m.sg. perfect form. A similar – yet not altogether identical – structure has been documented in the dialect of Najrān, just across the Saudi border.1 The linguistic process discussed here has resulted in three distinguishable patterns, namely faʿaltant, faʿaltint and faʿalhant, each typical of certain villages scattered around the Yemenite province of Ṣaʿdah and in Najrān in Saudi Arabia. Abbreviations: Localities in the Province of Ṣaʿdah AŠ Am Bṭ BW Ġr

Āl Šalīl Amlaḥ Baraṭ Bāni-ʿWēr Ġarīr

Ḥd Rz Sg Sn Nj

Ḥaydān Rāziḥ Sāgēn Sṓdān Najrān

1. The pattern faʿaltant The motive behind the development of the pattern faʿaltant apparently lies in the structured ambiguity in marking the pronominal subjects of the perfect, which results from the formal identity between the 1.sg. and the 2.m.sg. Following the omission of the historical final vowels, old faʿaltu and old faʿalta both result in faʿalt, which is now used for both persons. This applies to the great majority of modern Arabic dialects, many of which display a certain degree of                                                                                                                 1

The main ideas of this article were presented in the 9th conference of the Association Internationale de Dialectologie Arabe, held in March 2011 at the Università G. D’Annunzio, Pescara.

Agglutinated verb forms in the Northern province of Yemen

261

ambiguity in the perfect verbal paradigm: the form faʿalt is open to multiple analyses, i.e. there is more than one way to analyze the input data. The ambiguity of a linguistic form often triggers language change. Certain Arabic dialects avoid this kind of ambiguity by preserving the final vowels: the hallmark of many Mesopotamian dialects is the verbal form qәltu for the 1st person singular, which is clearly distinct from qәlt – “you (m.sg.) said” in the same dialects. In a similar manner, several Yemenite k-dialects have developed a morphological distinction between the first and second person. Some mark the first person with a final vowel, as qulku vs. qulk (“I said” vs. “you said”) in the dialect of Taʿizz (Prochazka 1974: 439)  while in other dialects it is the second person which is marked, as katabt vs. katabta (“I wrote“ vs. “you wrote“) in the dialects of Riṣābah and Ḥammām ʿAlī (Behnstedt 1985: map 69). In others the differentiation is based upon the inner vowels, with katubk vs. katabk (“I wrote” vs. “you wrote”) in Jiblih (Diem 1973: 100); or sarḥuk vs. sarḥik (“I went” vs. “you went”) in the Arabic of Jabal Rāziḥ (Behnstedt 1987: 133). The dialects of the Northern Province of Yemen and that of Najrān offer an essentially different mechanism to maintain the functionally important difference between the persons and disambiguate the formal identity: differentiation is achieved by marking the 2.m., i.e. by the addition and subsequent agglutination of the independent pronoun ant to the basic faʿalt form, thus making it unequivocal and transparent. The form faʿalt now denotes the 1.sg., while faʿaltant is used for the 2.m.sg. (one specific circumstance in which non-agglutinated forms are used for the 2.m.sg. is mentioned further below). Preservation of final vowels Mesopotamia Taʿizz Riṣābah 1.c.sg. qәltu

qulku

katabt

2.m.sg. qәlt

qulk

katabta

Distinction of inner vowels Jiblih

J. Rāziḥ

Agglutination Ṣaʿdah

1.c.sg. katubk

sarḥuk

1.c.sg. katabt

2.m.sg. katabk

sarḥik

2.m.sg. katabtant

Most interestingly, Ingham (1994: 23-4) mentions that in Najdi Arabic the independent pronoun ant is often postposed to the verb, as in kitabtih ant – “did

262

Ori Shachmon

you m. write it?”, and suggests that the reason may be that kitabtih could mean either “I wrote it” or “you m. wrote it”. However, in Najdi Arabic this ambiguity only occurs when the verb is followed by an object pronoun, while without it the 2.m.sg. yields kitabta, which is clearly distinguished from kitabt (“I wrote”). Following in the table below are the complete perfect paradigms of the different verb types used in the Yemenite dialects discussed. Note, that vowels in forms of the pattern fiʿíl vary in accordance with the quality of the adjacent consonants, as discussed in detail in Shachmon 2007: 92-97. Type a faʿál

Type i/u faʿíl

fiʿíl fiʿíl

fuʿíl

fuʿúl

1.c.sg.

katabt

jalist

širibt

kubirt

gurubt

2.m.sg.

katabtant

jalistant

širibtant

kubirtant

gurubtant

2.f.sg.

katabtī

jalistī

širibtī

kubirtī

gurubtī

3.m.sg.

katab

jalis

širib

kubir

gurub

3.f.sg.

katabat

jalsat

širbat

kubrat

gurbat

1.c.pl.

katabnā

jalisnā

širibnā

kubirnā

gurubnā

2.c.pl.

katabtū

jalistū

širibtū

kubirtū

gurubtū

3.c.pl.

katabū

jalsū

širbū

kubrū

gurbū

While the preservation of the final vowel in the above-mentioned Mesopotamian (viz. qәltu) and Yemeni dialects may be explained in terms of “paradigmatic resistance to sound change” (Blau 1985: 6), the case of faʿaltant is of a quite different nature. The synchronic layout of the North-Yemenite paradigm implies that the paradigmatic sound change did operate, resulting in the coalescence of *faʿaltu and *faʿalta into faʿalt, but it was then followed by the innovative introduction of a differentiating element, viz. agglutinated -ant. Through this cliticization one of two formally identical forms becomes “marked” and ambiguity is precluded. Note that the synchronic result of this cliticization in the dialects discussed is that the independent pronoun and the subject marker are now homophonous, viz. both ant.2

                                                                                                                2

The cliticization of pronouns to verbs may be seen as a cyclic change, which is potentially operative and then re-introduced when a sufficient motivation evolves.

Agglutinated verb forms in the Northern province of Yemen

263

Following are a few examples of faʿaltant forms in their natural context, as recorded from various villages in the Province of Ṣaʿdah:3 •

gām, gid alkull ḥāṣil. jō axwatih: rētant ḥad willā šuftant ḥad? gāl: walā rēt ḥad. mā gumt illā šuft ... kull šī ḥāṣil (Am) – He got up and everything was ready. His brothers came: Did you see anybody or notice anybody? He said: I saw no one at all. No sooner had I got up than I saw ... everything was ready.

•

ant aštarētant xuḏ̣rah fi assūg (Sn) - You bought vegetables in the market.

•

aḥzir ʿādak gid aġtamētant aw ʿādak bxēr4 (Ben-David 1999: 61) – I am looking to see whether you are still unconscious or already well again.

•

midri mā jā ʿalēh. gid hū šēṭaHb  5. mā jā lak? mā ʿalēk? mā sabbartant? mā fiʿiltant? gāl: sirt, gāl, ʿalā lgābarHb ḥagg ašŠabazī́ (Ḥd) – I don’t know what has occurred to him. He was [behaving like] a fool. [He said to him:] What is wrong with you? What have you done? What have you done? He said: I went, he said, to the grave of ašŠabazi6.

•

gāl: hā ̮smaʿ! albahmah ḥaggatī ḏ̣aʿīfah. in aʿjabatak šallētanthā w-in mā aʿjabatak xallīhā (Ben-David 1999: 63) – He said: Hey listen! My cow is weak, if you like it you [may] take it, and if you don’t like it – leave it.

•

wāḥid sāfar la-TelavivIHb kān bidduh 7 yištarī́ ótoIHb... sār yixabbír la-ṣāḥbuh. gā́l luh: sirtant Talabīb? rētant almakīnah? lēš mā aštarētanthā? gāl: innhā ṣ̮ ġáyyirah (Rz) – A person went to Tel-Aviv, he wanted to buy a car... He told his friend about it. He [the friend] said to him: Did you go to Tel Aviv? Did you see the car? Why haven’t you bought it? He said: It was [too] small.

                                                                                                                3

My examples for the agglutinated forms were recorded from Yemenite Jews who immigrated from the province of Ṣaʿdah to Israel in the 1950s, 1960s and 1990s. For more details see §4 below.

4

The adjective bxēr, derived from the combination of b + xēr, is no longer separable and even has a plural form bxērīn, as in the example mā fīnā marīḏ̣ kullanā bxērīn – “None of us is sick, we are all well” (Ben-David 2010, vol. 1: 331). In addition, several verbal forms are formed from the denominative roots √bxr and √bxyr (Piamenta 1990: 21).

5

Hebrew, Aramaic and Judeo-Arabic components are interwoven throughout the texts I recorded from Yemenite Jews. Hebrew components which were part of the language already in Yemen are labelled Hb, and should be differentiated from forms and roots which were borrowed after the immigration to Israel (marked IHb). For more details see Shachmon 2013 and the reference given there.  

6

Rabbi Shalom Shabazi (d. 1720) was one of the most famous Yemenite Jewish poets and his grave in Taʿizz is a place of pilgrimage for Jews and Muslims alike.

7

This example was recorded in the speech of a Jew from Jabal Rāziḥ, who used the pronominal suffix -uh as opposed to -ih which is more regular in the dialects of the Northern Yemenite Jews.  

264

Ori Shachmon

The forms šallētanthā and aštarētanthā of the last two examples show that the process of agglutination is actually complete: the former independent pronoun, viz. the ant segment, has become an integral part of the conjugated verb, is no longer separable and remains attached to the verb even when a pronominal suffix follows it. The same is true of the clitic prepositions, as in the examples gúltantilhā (“you said to her”) or katabtantlī máktūb (“you wrote a letter to me”). There is, however, one circumstance in which the motivation for this kind of agglutination does not evolve. The following is an extract from a story about a young boy who has been possessed by jinn. As he comes back to his senses the worried family tries to understand his weird behaviour, using the non-agglutinated verb fajaʿtanā - “you have frightened us”:8 kān nugullih: Saʿd! ismih Saʿd, ism alwald Saʿd. kān yigullih abī: Saʿd wēš mā-lak? wēš mālak yā-bā? hā gūm itgambas, wēš mā-lak? gāl: wēš mā-lī yā ʿamm Sālim? wēš sawwēt? gāl: wēš mā-lak? wēš mā-lak? yā waladī ant fajaʿtanā wēš mā-lak, wēš bēn yūjʿak? gāl: mā-li ḥājah. māna dārī šī. anā rāgid mā sawwēt šī (Sg) We were calling to him: Saʿd! His name was Saʿd, the name of the boy was Saʿd. My father was saying to him: What is up with you? What is up with you, my dear? Get up, sit upright, What is up with you? He [the boy] said: What is up with me, uncle Sālim, what have I done? He [the father] said: What is up with you? what is up with you? My boy, you have frightened us, what is up with you? What is hurting you? He said: Nothing is wrong with me, I have no idea. I was sleeping, I did not do anything.

In a form like fajaʿtanā (or, in comparable examples as ʿaraftanī – “you m.sg. recognized me”; jibtanā – “you m.sg brought us”) it is rather unlikely that both the verb and the pronoun would denote the 1st person. As in other Semitic languages, in Yemenite Arabic object suffixes of the first and second person are only very rarely attached to verbs of the same person. Ambiguity is precluded in such cases by morphology, and the speakers intuitively interpret the accurate meaning of “you m.sg. did X to me/us”. 2. The pattern faʿaltint The informant from whom I recorded the text in §5 below consistently uses the pattern faʿaltint rather than faʿaltant. He originates from the village of Āl-Šalīl,                                                                                                                 8

 

Most importantly, the a which precedes the attached 1.pl. suffixed pronoun -nā is by no means the vowel of the old 2.m. suffix -ta, but rather an anaptyctic inserted between any form ending in v̄C or vCC and a pronoun opening with a consonant, e.g., bēt + nā > bētanā (“our house”) or ḥagg + kum > ḥaggakum (“yours”), as characteristic for many Arabian dialects.

Agglutinated verb forms in the Northern province of Yemen

265

north of the city of Ṣaʿdah in the Northern province of Yemen (see Behnstedt 1987, map 2: Gegend von Ṣaʿdah). Similar forms have been documented in the nearby village of imṬalḥ (Behnstedt 1987: 28). The following extract from his speech shows several examples of the faʿaltint pattern: gāl: ṣihrak, nasabak, ištkā́-bak, gāl innak daffartint maratak w-xabáṭṭinthā ... gult: bā-addīĺak maṯnī lā rayyaʿtintlī hōn. lā jītint mʿī mānā bā-addīlak šī. (paragraphs 3-4) He said: Your brother-in-law, your relative, has informed on you and said that you had expelled your wife and had beaten her ... I said: I will give you double [payment] if you wait for me here. If you come with me I will not give you anything.

The pattern faʿaltint is rarer than faʿaltant and matches the shift of a > i in closed syllables which also occurs in some of the independent pronouns. For example: iḥnā and intū were documented in Rāziḥ and Sāgēn, as well as in ĀlŠalīl itself, as opposed to aḥnā and antū which generally prevail in this area. I recorded int and intī for the second person singular from the Jews of Āl-Šalīl, and Behnstedt (1987: 64) had already documented them in imṬalḥ and Ṣaḥn, all in very close proximity. The distribution of the various pronouns is given in the table below:

Am

sg. 1.c.

2.m.

2.f.

3.m.

3.f.

pl. 1.c.

2.c.

3.c.

anā

ant

antī

hū

hī

aḥnā

antū

him

AŠ

anā

int

inti

hū

hī

iḥnā

intū

him

Bṭ

anā

ant

antī

hū

hī

aḥnā

antū

him

BW

anā

ant

antī

hū

hī

aḥnā

antū

him

Ġr

anā

ant

antī

hū

hī

aḥnā

antū

him

Ḥd

anā

ant

antī

hū

hī

iḥnā

antū

him

Nj

anā

ant

antī

hū

hī

aḥnā

antū~intū

him

Rz

anā

ant

antī

hū

hī

iḥnā

antū

him

Sd

anā

ant

antī

hū

hī

aḥnā

antū

him

Sg

anā

ant

antī

hū

hī

iḥnā

intū~antū

him

In addition, several verbal forms in the dialect of Āl-Šalīl exemplify the same shift of a > i in closed syllables, as itḥāmaṭ – “I get upset” instead of atḥāmaṭ which prevails in this area; ištaġal – “I work”, instead of aštaġal, etc. Note, however, that the shift does not occur in other morphological categories: •

kull mā gallī min kalmah antfaxt ... bēn itḥāmaṭ ʿalēh – With every word he said to me I swelled up more and more, I got upset with him.

•

kān ištarī ḥaṭab min assūg – I would buy firewood from the market.

266

Ori Shachmon

•

gid anā́ bēn íštall min algāʿ min alḥámaṭah – I almost picked myself up off the ground with nerves.

•

limah? ʿalašān anā bēn ištaġal fuḏ̣̱ḏ̣ah – Why? Because I used to work with silver (i.e. as a silversmith).

•

ṭaliʿt, gid ana bēn intfuḏ̣ – I went up, I was shaking all over.

3. The pattern faʿalhant Side by side with faʿaltant and faʿaltint, a third pattern of agglutination may also be heard in this area, namely faʿalhant. The latter appears in the speech of informants from Najrān, as already documented by Th. Prochazka in his survey of Saudi dialects (1988: 12, 24, 28, 77). In fact, forms of the pattern faʿalhant are known locally as a hallmark of Najrāni Arabic. For example, a web forum providing a “conversation manual” for the Arabic dialect of Najrān instructs learners to add (sic!) hant to the perfect verbal form of the 2.m.sg., i.e. ant katabhant instead of ant katabt (“you m.sg. wrote”)9. Within the province of Ṣaʿdah, I heard faʿalhant forms from speakers hailing from Wādi Amlaḥ, yet it is not in use among all Amlaḥis. According to my inquiries those speakers who use the faʿalhant pattern had immigrated from Najrān southwards in the 1930s and 1940s. It is noteworthy, that Behnstedt (1987: 27) documented similar forms in Dammāj, south-east of Ṣaʿdah. Following are a few examples: •

katabhantlī waragah – You wrote me a note

•

gid sirhant albēt? – Have you gone home?

•

aštarēhantlī ḥājātī – Have you bought the goods (lit. my stuff) for me?

•

ant nsīhantanī – You forgot me!

At first glance, the pattern faʿalhant looks as if it is based on the 3.m.sg. verbal form, viz. faʿal + hant > faʿalhant. This, however, cannot be so. Forms of second or third weak roots indicate that faʿalhant is actually based upon the 2.m.sg., as expected: the form sirt (not: sār) must have been the starting point of sirhant (“you m.sg. went”), and rēt (not: rā) is definitely where rēhant (“you m.sg. saw”) began. Accordingly, it is clear that the pattern faʿalhant is structurally identical with faʿaltant except that it exhibits an h in place of the t of faʿalt. Previous attempts to explain its formation assumed the replacement of the original t of faʿalt by the independent pronoun ʾant, followed by a replacement of the hamza by h                                                                                                                 9

Taken from: http://forum.te3p.com/379953.html.

Agglutinated verb forms in the Northern province of Yemen

267

(Behnstedt 1987: 27-8; Prochazka 1988: 12, 24; Behnstedt-Woidich 2005: 143).10 Yet, a series of simple replacements does not seem to sufficiently explain these forms. Such replacements certainly did not take place in faʿaltant and faʿaltint, where t of faʿalt is clearly evident. The evolution of the faʿalhant pattern must have involved the agglutination of the independent pronoun ant to the conjugated 2.m.sg. verb faʿalt, without replacing the old -t. This stage is plain to see in the patterns faʿaltant or faʿaltint. At this point, the original pronominal -t was apparently felt redundant, since both -t and ant denote the 2nd person. The repetition of t in katabt and in ant is more than just a sequence of two similar sounds, it is pleonastic in form and in function – each of the t-sounds serves to indicate the 2.m.sg. Consequently, the first t, no longer serving any morphological purpose, became liable to reduction and eventually dissimilated to h. By all accounts, the divergence between faʿaltant, faʿaltint and faʿalhant does not go back to different starting points, but reflects successive developments from a single origin. In some dialects, an agglutination of faʿalt + ant further developed by dissimilation to become faʿalhant. 4. Geographical and Communal Distribution In Die Dialekte der Gegend von Ṣaʿdah Peter Behnstedt briefly mentions agglutinated forms of the type discussed, yet these are – according to his findings – in marginal use only (Behnstedt 1987: 27). Since the book deals primarily with the local Muslim population, one may infer that the agglutinated forms are known to the local Muslim population, though they are secondary to the more prevailing faʿalt pattern. On the other hand, the agglutinated forms are – without any doubt – the predominant forms in the speech of Jews originating in this very same province. My examples for the agglutinated verbal patterns were all recorded from Yemenite Jews who immigrated to Israel. Some of them left Yemen around 1950 in the famed “Operation On the Wings of Eagles”11; others stayed in Yemen until the early 1960s, and a few hundreds were brought to Israel in the mid1990s. The last few Jewish families who remained in the Northern province into the 21st century had to leave during the violent events which have been taking place in North Yemen since the early 2000s.                                                                                                                 10 Kaye (1990 : 86) even tried to connect -hant with Persian hast.   11 The immigration of about 50,000 Yemenite Jews to Israel between December 1948 and late 1951.

268

Ori Shachmon

In Israel, the Yemenite immigrants often choose to live among their fellow Yemenis and thus maintain their mother-tongue in active use. The old generation still preserves many of its former linguistic habits, including the unique agglutinated patterns discussed in this article. These forms are actually an earcatching indicator by which immigrants from the northern province of Yemen are easily identifiable among other Yemenites, even in the midst of a crowded social gathering. But while Yemenite Arabic is still in use among the immigrants themselves, the young generation naturally switches to Israeli Hebrew, and – just like other immigrant languages – Jewish Yemenite Arabic is gradually dying out in Israel. The use of agglutinated verbal forms of the pattern faʿaltant thus seems to be disappearing. These forms are apparently marginal in Muslim speech around Ṣaʿdah, and the Jews who used to use them in Yemen are now mostly in Israel and will eventually stop using Arabic altogether. At the same time, in the dialect of Najrān faʿalhant forms are common in regular Muslim speech, and seem to have a better chance of surviving. 5. Sample text I recorded the following text in 1996 in the small town of Bnei ʿAyish in central Israel. The speaker is a Yemenite Jew who immigrated to Israel in 1962. He was born around 1930 in the village of Āl-Šalīl. He is married to an immigrant from Najrān, and they speak Yemenite Arabic at home. The neighbours in their close vicinity are also of North Yemenite origin, and so the dialect is fairly well preserved in their daily speech. In this story the informant gives an illustration of his firm and respected position among government officials in Yemen. In a few points he addresses me directly (in the 2.f.sg.) to interpret local terms and concepts which he had not been convinced that I knew. Among the Hostages in a Yemenite Jail 1. gulnā: hā bā t̮ záwwaj. gāl wāḥid: iddīnī uxtak w-anā bā d̮ dīk uxtī. gult: wálā yhimmak. addēt uxtī w-hū addānī uxtih. anā mā sāʿadatni uxtih. hināk diktatūraIHb: ilā-mā tʿaddat wāḥid almarah bā-yuxbuṭhā willā yiḥbishā, ilā mā tʿaddatih. We said: I should get married. Someone said: Give me your sister and I will give you my sister. I said to him: No problem. I gave my sister and he gave me his sister. As to myself, his sister didn’t please me. It’s dictatorship there. When

Agglutinated verb forms in the Northern province of Yemen

269

a woman disobeys a man he beats her or imprisons her, that is if she disobeys him. 2. tšā́jar niḥnā12 wiyyāhim. w-him iṯnēnā w-anā́ mā mʿī šī, fágīr, wāḥid waḥadih, lā mʿī lā ʾax walā mʿī ʾab walā mʿī ayy ḥājah mā anā ̮llā waḥdī. kān mʿī maṣnaʿ ḥagg alʿarab w-kān aṣāḥibhim. alḥkūmah - lā bih ḥākim wilā ʿaskarī willā ... kān anā wiyyāhum ṣadīg. wēš immā gult lhim gālu ... We argued, we and them. They were two while I had nothing, a poor guy. One by himself. I did not have a brother, I did not have a father, I did not have anything. I was all by myself. I had a workshop belonging to the Arabs and I used to socialize with them. The authorities ... Say there was a governor or a soldier, I was friends with them. Whatever I would say to them they would say [yes] ... 3. yōm wāḥid jīt w-gid-anā́ jāy gim-mʿī13 dijājah w-gim-mʿī́ galīl ḥabb w-gid-anā́ jāy. daxált Ṣaʿdah, gillī14 xārij Ṣaʿdah, ištáġalt. ilʿárab addōlī ḥabb w-addōlī dijājah gid-anā́ ḏạ̄ wī albēt. ilā w-ʿaskarī jāw15 yōm alxamīs. gāl: ḥabs. gult: lēš? gāl: ḥabs yā malʿūn. gult: lēš? gāl: ṣihrak, nasabak ištkā́-bak gāl innak dáffartint maratak w-xabáṭṭinthā. One day I came, and when I came I had with me a hen, and I had with me a little grain, and so I came. I entered Ṣaʿdah, I had been out of Ṣaʿdah, working. The Arabs had given me grain, given me a hen, and I was coming back home. Suddenly, there came a soldier, it happened to be Thursday. He said: Jail. I said: Why? He said: Jail, you cursed one. I said: Why? He said: Your brother-in-law, your relative, has informed on you and said that you had expelled your wife and had beaten her.                                                                                                                 12 The 1.c.pl. independent pronoun prevailing in these dialects is iḥnā~aḥnā (see §2 above), yet niḥnā~naḥnā often occurs following verbs denoting reciprocity. In such cases the verb appears to be in the singular, which implies that the n in niḥnā~ naḥnā is most probably not a remnant of OA naḥnu but rather of the 1.pl. subject pronoun: *tšājarnā + iḥnā > tšājar niḥnā. Comparable examples are tgātal niḥnā wiyyāh (Sg) – “we [and him] had a quarrel” or atḥōramHb naḥnā wiyyāh (Ben-David 1999) – “we and him cursed each other”. 13 Assimilation of gid + mʿī > gim mʿī. 14 Plausibly gid + lī > gillī. Compare gad + linā > gállinā in Behnstedt 1987: 171 and n.16. 15 The idiom jāw contains the verb jā, which is fully fossilized in this construction and may no longer be inflected for gender and number, followed by the conjunction w-. In the dialect discussed the two elements function as one unit denoting “it so happened that ...” or “it turned out that…”, referring to a continuous state in the past. jāw is frequently used as a literary connecting device, often opening a story or an anecdote, bringing a sense of coincidence or chance into the narrative.

270

Ori Shachmon

4. gult: walā yihimmak, xallīnī anázzil addijājah albēt w-anazzil alḥabb ḥaggī albēt wanā bā ̮ṭlaʿ alḥabs. rayyiʿlī hinā wirjaʿ bēn agūl bā-addī́lak maṯnī, ʿalašān lā yijī mʿī. gāl: kwayyis. arayyiʿlak hōn?16 gult: rayyiʿlī hōn. gult: bā-addīĺak maṯnī lā rayyaʿtintlī hōn. lā jītint mʿī mānā bā-addīlak šī. gāl: wálā yihimmak. I said: No problem. Let me bring the hen home and bring my grain home and I will go to jail. Wait for me here, and then – so I said – I will give you double. [I said this] in order that he doesn't come with me. He said: Okay. Shall I wait for you here? I said: Wait for me here. I said: I will give you double if you wait for me here. If you come with me I will not give you anything. He said: No problem. 5. mā billā názalt ḥaṭṭēt alḏā w-šallēt nafsī w-ṭaliʿt ... alḥkūmah ... aḏill min álḥākim ḏā! alḥākim alkabīr, ṭaliʿt ilā ʿandih. waṣilt ... hā kān mʿī hināk ṣadīg. waṣilt gid hum almaġrib. gult: itballaġ bī la-ʿand annāḏ̣rah. gāl: lēš? xabbart kaḏā́ w-kaḏā́ w-kaḏā́ wkaḏā́. I simply went, put those things away and took myself and went ... the authorities ... I was afraid of that governor! [He was] the governor general. I went up to him. I arrived, I had a friend there. I arrived while they were at the evening prayer. I said: Let me be received by the arbitrator. He said: Why? I said, I told him so and so and such and such. 6. gāl: wálā yhimmak, wálā yhimmak. itbállaġ bī gāl: yikūn, iṭlaʿ! nasabih w-abū almarah – alḥabs. yiṭallʿūhum alḥabs. ṭállaʿūhum alḥabs ʿindī, ṭaliʿnā kullanā alḥabs, daxalnā alḥabs. hā w-makānanā. gāl ... kull mā yigaddimu škiyyah la-ʿand alḥākim yigūl yitrāḏ̣ā alġarīmēn. He said: No problem, no problem. I was received, he [the arbitrator] said: Let it be, come up! His brother-in-law and the wife's father shall go to jail. They are to be taken to jail. They took them to jail along with me. We all went to jail. We entered the jail. That is our situation. He said: Whenever a complaint is brought before the arbitrator he says: Let the two disputants come to terms. 7. ánā kān mʿī ṣadīg alʿaskar, kān yiddōlī ukul w-yiddōlī ḏā, mikayyif mistarīḥ, w-algāt w-almadāʿah w-alkiyāfah. w-mʿī ḥarr ʿindhim addōlī waḥadī. w-him ḥaṭṭūhum bēn arrahāyin. w-bēn arrahāyin hināk kān hū makān ḥamīr w-algámil w-algummal. yidaxxlūhum hināk. makān zift.                                                                                                                 16 The forms kwayyis and hōn are not part of the original dialect. The informant is imitating the soldier, using forms which sound “different”, though he may have learnt the “different” forms from his Arab friends in Israel and not from speakers of other dialects in Yemen.

Agglutinated verb forms in the Northern province of Yemen

271

Me, I had a friend among the soldiers. The were giving me food, giving me qāt. I was having fun, enjoying myself, [chewing] qāt, [smoking] the water-pipe, having fun. I had a cell there, they gave it to me for myself only. As for them, they put them among the hostages.17 Among the hostages there, it was a place of asses, of lice and fleas. They put them in there. A terrible place. 8. him gib ̮bā-yimūtū min aljūʿ min alʿaṭaš w-min alkull, min arrahāyin – mantī dāri wēš arrahāyin?antī dārī wēš arrahāyin? anā bā ̮xabbiriš: ilā bih šēx w-mʿih xḗrāt warā́ ḏ̣ahrih bēn yiddī rahīnah, yiddī ibnih w-yiṭallʿūh hināk, rahīn. hū yiddī rahīnah hināk w-yugʿudū hināk. They were nearly dying from hunger and thirst and everything, from the hostages – don’t you know what “hostages” are? Do you know what “hostages” are? I shall explain this to you: if there is a tribal chief who has many [tribesmen] backing him, he gives a hostage, he gives his son and they take him there [i.e. to jail] as a hostage. He gives a hostage there and they [i.e. the hostages] stay there. 9. w-anā kān mʿī makān, mikayyif argud w-ākul w-aʿammir miṯilḥīn anā́ hinā fi-bētī. hā kān fīnī gēd w-alʿaskarī́ bēn yixallīlī wāḥid máftūḥ. mā billā kamā́ b-wuddī mā billā bēn abʿíd alḏā w-aḥuṭṭih fi rrijil aṯṯānī w-asīr. And me, I had a my own place, I was having fun, sleeping and eating and smoking [lit. filling the water-pipe] like when I were here at home. Now, I had chains on me, and the soldier leaves one of them open for me, so that whenever I want I just remove it and put it on the other leg, and walk. 10. w-him migayyadīn. avalIHb kull gēd bījī hū miṯil ḏā mā hū miṯil ḥaggōn hinā, ḥadīd! bēn yisawwōh alyihūd. kān yisawwōh alyihūd, kānhim ḥáddādīn alyihūd. kull wāḥid miṯil yaddī bēn yikawwin arríjlēn. lā bā-tsīri galīlah w-rijliš gid hī malāyah dam. lā kān yisáwwō ʿalēhā šalāyil. As to them, they were chained. Each chain was something like that,18 it was not like those of here. Iron! The Jews used to make them. They were blacksmiths, the Jews. Each one is [as thick] as my hand, it wounds the legs. If you walk for just a little your leg will already be full of blood. They didn't [even] use to put rags on it.                                                                                                                 17 In order to guarantee the loyalty of tribes, Yemenite Imāms used to incarcerate one or more near relatives of the chiefs of important tribes, often young boys, keeping them as rahāyin or hostages in special prison schools. See Wenner 1967: 78. This custom is also mentioned in the memoirs of Abraham Ovadiah 1985: 30. 18 The informant used his hands to demonstrate the thickness of the chains.

272

Ori Shachmon

11. jā wāḥid šēx min hināk: ṭaliʿ w-kān hū ṣadīg hū wiyyāh, hū wiyyah, nasabī. jā gāl: hā bā-asidd bēnakum. gult: lā yisidd bēnanā illā wuddih yiṭallig uxtī w-anā bā-aṭallig uxtih. willā mā nuxrij kaḏā gult mā billā makānanā hinā. Some sheikh came from there. He came, and he was a friend of his, of his, that is of my relative. He came and said: Hey, I shall reconcile between you. I said: He will not reconcile between us unless he is willing to divorce my sister and I divorce his sister. Otherwise we don't come out, I said, we are staying here. 12. yā ḏāk. yā ḏāk. gult: mā billā makānanā. urgum ilā wuddak innih yiṭallig uxtī. whū yiṭállighā áwwal, hū gabil, wírjaʿ anā aṭallig uxtih. wirjaʿ sawwā́ bēnanā - ḏā ʿandiš arragim. ṭaliʿnā alḥākim inn iḥnā istarḏ̣ēnā w-kull wāḥid yiṭlig aṯṯānī. áṭragnā. wbaʿdēn, baʿdēn xarajnā, hū ṭallag uxti, hū ṭallag uxtī w-anā́ ṭallagt uxtih. [He tried convincing me:] Hey there, hey there. I said: we are staying here. Write down [a settlement], if you want, that he should divorce my sister. And he should divorce her first. He should be first, and then I will divorce his sister. Then he arbitrated between us - here is the note in front of you.19 We went up to the arbitrator [saying] that we have come to terms, and each one will divorce [the sister of] the other. We left. Afterwards we went out. He divorced my sister, he divorced my sister and I divorced his sister.

Bibliography Behnstedt, Peter. 1987. Die Dialekte der Gegend von Ṣaʿdah (Nord Yemen). Wiesbaden: Harrassowitz. Behnstedt, Peter and Manfred Woidich. 2005. Arabische Dialektgeographie. Leiden: Brill. Ben-David, Aharon. 1999. Hawīdah ve-Dāwid mi-Tsfon Teman. Rehovot: Ahavat Teman. — 2010. Sefer ha-maʿaśim. 2 vols. Rehovot: Ahavat Teman. Blau, Joshua. 1985. “On Some Arabic Dialectal Features Paralleled by Hebrew and Aramaic.” The Jewish Quarterly Review, New Series, Vol. 76, No. 1: 5-12. Diem, Werner. 1973. Skizzen jemenitischer Dialekte. Beirut: In Kommission bei F. Steiner, Wiesbaden. Ingham, Bruce. 1994. Najdi Arabic: Central Arabian. Amsterdam and Philadelphia: J. Benjamins. Kaye, Alan S. 1990. Review of Prochazka 1988. Zeitschrift für arabische Linguistik 21: 8388. Ovadiah, Abraham. 1985. Netivot Teman ve-Ziyyon. Tel-Aviv: Afikim.

                                                                                                                19 The informant illustrates writing a settlement on a piece of paper.  

Agglutinated verb forms in the Northern province of Yemen

273

Prochazka, Theodore Jr. 1974. “The perfect tense ending -k in the spoken Arabic of Taʿizz”, Bulletin of the School of Oriental and African Studies 37, No.2: 439-442. — 1988. Saudi Arabian Dialects. London: Kegan Paul International. Shachmon, Ori. 2007. Arabic Dialects of the Jews of North Yemen. Thesis submitted for the degree of "Doctor of Philosophy". The Hebrew University of Jerusalem. — 2013. “Judeo-Arabic, Yemen, Hebrew Component.” Encyclopedia of Hebrew Language and Linguistics. Edited by Geoffrey Khan. Vol. 2, pp. 402-406. Leiden: Brill. — 2014. “ibn aššarīfah vs. ibn aljāriyah, The Hebrew component as a polemic device.” The Jewish Quarterly Review 104/1: 144-166. Wenner, Manfred W. 1967. Modern Yemen – 1918–1966. Baltimore: Johns Hopkins Press.

Qusṭā b. Lūqā. On protection against rheum and catarrhs that occur in the winter. Edition and English translation of the Arabic text preserved in the manuscript Ayasofya 3724 Lena Ambjörn, Lund University

Introduction Qusṭā b. Lūqā al-Baʿlabakkī al-Yūnānī is mentioned in several bio-bibliographical sources,1 but the information is scarce. It is clear that he was active around 860 CE, mainly in Baghdad and Samarra, in the prestigious and well-paid circle of scholars associated with the Abbasid court.2 The name indicates that he came from Baalbek (Baʿlabakk) and was of Greek origin. He knew Greek, Syriac and Arabic, and was one of the key figures in the translation movement which was initiated by the Abbasid caliphs in order to get direct access to Greek sources and reached its peak in the 9th century.3 Among his colleagues were the slightly older Ḥunayn b. Isḥāq (808–873/877), who translated Galen’s books into Syriac and Arabic4, and the mathematician Thābit b. Qurra (834–904). Qusṭā translated Greek works on geometry and mechanics for influential commissioners, one of whom was Prince Aḥmad, the son of the caliph al-Muʿtasim.5 Apart from his translations, Qusṭā produced more than fifty books and articles, mainly on medicine6, but having been trained in the quadrivium disciplines, he was also skilled in astronomy, mathematics, logic and musicology. The medical texts attributed to Qusṭā deal with a wide range of topics: there are treatises on basic physiology according to the Galenic system of humoral medicine, and on particular diseases such as gout, numbness, dental nerve dysfunction and common 1

Ibn Ǧulǧul, Ṭabaqāt, 76; Ibn an-Nadīm, Fihrist, p. 295; Ibn al-Qifṭī, Taʾrīḫ al-Ḥukamāʾ, pp. 262f.; IAU, vol. I pp. 244f. and II p. 166.

2

See Correspondance pp. 28f. and p. 117, § 126.

3

For a discussion of the translation movement and its motives, see Gutas, Greek thought.

4

For Ḥunayn’s own account of the translations, see Mā turǧima, ed. Bergsträsser.

5

For Qusṭā’s translations, see Gabrieli, Nota, and Gutas, Greek thought.

6

For Qusṭā’s medical works, see GAS III pp. 270–274; for works on mathematics and astronomy, GAS V pp. 285ff., and VI pp. 180ff. Some of the identified MSS are not available since the collections are dispersed or lost.

Qusṭā b. Lūqā. On protection against rheum and catarrhs

275

cold. He also wrote on sexual potency, the transmission of diseases, the causes of longevity, moral character, and an important work on the difference between pneuma, the agent of sensation and motion (Ar. rūḥ) and the soul (nafs). He produced a health-guide for pilgrims going to Mekka, and an abbreviated Arabic version of one of Galen’s books on phlebotomy. Of his medical works, approximately a fourth is now available as text-critical editions, with translations and more or less extensive commentaries. He explicitly states that his patrons expected him to consult the works of the ancient authorities and relate the information in a concise, systematic and comprehensible way. To generate original and new ideas was not his task. Qusṭā was not mainly a clinically active physician, but rather a scholar who concentrated on theoretical reasoning, following the scholastic tradition of Late Antiquity. Qusṭā was regarded as an authority on medical as well as other scientific issues not only during his lifetime, but for several centuries after that. He is quoted in later medical works, such as arRāzī’s (d. 925) Kitāb al-Ḥāwī fi ṭ-Ṭibb (Lat. Liber Continens) and Ibn al-Jazzār’s Zād al-Musāfir, also from the 10th century, and in Biruni’s (d. 1048) pharmacological book Kitāb aṣ-Ṣaidana fi ṭ-Ṭibb. Some of Qusṭā’s own works were also translated into Latin.7 Qusṭā b. Lūqā’s short treatise On protection against rheum and catarrhs that occur in the winter is preserved in the manuscript MS Ayasofya 3724, fols. 99a– 103b. That the Abbasid caliph al-Mutawakkil (d. 861) was affected by rheum from the smell of roses is discussed in the text, but who actually commissioned the work is not clear. As in all medieaval medical works, the theoretical basis is that of Hippocratic-Galenic humoral medicine. Thus, an imbalance of the four cardinal qualities (hot, cold, dry and moist) represented in the body by the four humours (black and yellow bile, blood and phleghm) is regarded as the cause of all disorders. The therapeutic principle is to treat the imbalance with the opposite: overheating is treated with remedies that were thought to have cooling properties, overfilling is treated by evacuation, e.g. laxation, vomiting or phlebotomy. Diet and regimen in a wide sense, including not only food and drink, but all habits of daily life such as sleep, baths, exercise, work and pleasure, are of great importance for both prophylaxis and treatment.8 In the treatise, almost eighty simple or complex medicaments are mentioned.

7

See Wilcox.

8

See e.g. Temkin, Galenism, and Siegel, Galen’s System of Physiology and Medicine.

276

Lena Ambjörn

In what follows, an edition of the Arabic text and an English translation of Qusṭā b. Lūqā’s treatise on rheum and catarrhs is presented for the very first time. The edition is based on the only extant complete manuscript, MS Ayasofya 3724, fols. 99a – 103b. The text has been divided into paragraphs by me, since this underlines the clear structure of the treatise and makes it easier to refer to specific passages. Punctuation in the Arabic text has been added by me. The copyist’s frequent use of final ʾalif mamdūda has been changed into ʾalif maqṣūra without annotation in the apparatus. Additions to the text are marked by . When making the translation, I have tried to produce a readable English while at the same time trying to stay close to the Arabic source. Technical terms have been carefully considered, and each technical term has, as far as possible, been trans-lated using one and the same English term throughout the treatise.9 Due to the problems of determining what they actually stand for, names of weights and measures have not been translated.10 Arabic, English and Latin names of all materia medica mentioned in the treatise, simple as well as compound prepara-tions, are found in separate lists below. Text

‫كتاب قسطا بن لوقا اليوناني في التحّرز من الزكام والنزالت التي تعرض في الشتاء‬ ‫بسم اهلل الرحمن الرحيم‬ ‫ا‬F‫رأس وم‬F‫ة ال‬F‫وب‬F‫ثرة رط‬F‫ن ك‬F‫ن ع‬F‫كائ‬F‫ام ال‬F‫زك‬F‫دوث ال‬F‫ن ح‬F‫ه م‬F‫ّرز ب‬F‫ا يتح‬F‫ير م‬F‫دب‬F‫دا ت‬F‫ر على ه‬F‫ذك‬F‫ وال‬.١ ‫ق‬F‫ري‬F‫بدن وط‬F‫ود في ال‬F‫رار األس‬F‫ة على امل‬F‫دال‬F‫ات ال‬F‫عالم‬F‫نف وال‬F‫ال ع‬F‫رأس ب‬F‫ن ال‬F‫ذب م‬F‫عدة ويج‬F‫ينقي امل‬ ‫تسخني‬F F ‫تجاوز في ال‬F F ‫ذي ينبغي أن ال ي‬F F ‫ّد ال‬F F ‫صدر والح‬F F ‫رأس وال‬F F ‫ه ال‬F F ‫قوى ب‬F F ‫ا ي‬F F ‫نه وم‬F F ‫عالج م‬F F ‫ال‬ ‫اء‬F ‫ه إن ش‬F ‫ا يكتفى ب‬F ‫عاني م‬F ‫ذه امل‬F ‫ن ه‬F ‫د م‬F ‫ل واح‬F ‫ثبت في ك‬F ‫ا م‬F ‫تجفيف وأن‬F ‫يب وال‬F ‫ترط‬F ‫د وال‬F ‫تبري‬F ‫وال‬ .‫اهلل‬ ‫ن‬ ّ ‫يه إ‬F‫ول ف‬F‫ أق‬،‫رأس‬F‫ة ال‬F‫وب‬F‫ثرة رط‬F‫ن ك‬F‫ن م‬F‫كائ‬F‫ام ال‬F‫زك‬F‫دوث ال‬F‫ن ح‬F‫ه م‬F‫ّرز ب‬F‫ذي يتح‬F‫ير ال‬F‫تدب‬F‫ا ال‬F‫أم‬F‫ ف‬.٢ .‫ وطبيعية وعرضية معا‬،‫رطوبة الدماغ ال تخلو من أن تكون طبيعية أو عرضية‬

9 Glossaries of technical terms, with identification of the corresponding Greek terms, have been prepared for several texts and translations. For a list of such glossaries, see SavageSmith, “Sources and Procedures”. Pharmaceutical preparations, such as pills, troches, electuaries, plasters, ointments, oils, gargles, chewing-gums etc., are accounted for in Fellmann, Qalānisī.

10 For the problematic variations of weights and measures, over time and geographically, see Hinz.

‫‪277‬‬

‫‪Qusṭā b. Lūqā. On protection against rheum and catarrhs‬‬

‫‪ .٣‬ف‪F‬أم‪F‬ا ال‪F‬رط‪F‬وب‪F‬ة ال‪F‬طبيعية‪ ،‬ف‪F‬دل‪F‬يلها أن ي‪F‬كون ش‪F‬عر ال‪F‬رأس س‪F‬بطا وال ي‪F‬عرض ل‪F‬ه الصلع وال ي‪F‬خفّ‬ ‫ال ‪F F‬شعر ع ‪F F‬ن ال ‪F F‬ياف ‪F F‬وخ‪ ،‬وأن ي ‪F F‬كون االن ‪F F‬سان ك ‪F F‬ثير ال ‪F F‬نوم ق ‪F F‬ليل االح ‪F F‬تمال للسه ‪F F‬ر وأن ي ‪F F‬عرض ل ‪F F‬ه‬ ‫النسيان كثيرا‪ ،‬و إذا نام يثقل نومه ويعسر انتباهه‪.‬‬ ‫‪ .٤‬وأم‪F F‬ا ال‪F F‬رط‪F F‬وب‪F F‬ة ال‪F F‬عرض‪F F‬ية‪ ،‬ف‪F F‬دل‪F F‬يلها أن ي‪F F‬كون االس‪F F‬تنشاق ك‪F F‬ثيرا وال‪F F‬دم‪F F‬وع م‪F F‬ن ال‪F F‬عني سه‪F F‬لة‬ ‫االن‪F F‬كساب س‪F F‬ري‪F F‬عة وأن يس‪F F‬رع ال‪F F‬زك‪F F‬ام إلى االن‪F F‬سان ع‪F F‬ن االس‪F F‬باب التي ت‪F F‬ول‪F F‬د ال‪F F‬رط‪F F‬وب‪F F‬ات م‪F F‬ثل‬ ‫|‪|99b‬فصل الشتاء والتدبير املرطب باملطعم واملشرب‪.‬‬ ‫‪ .٥‬وأم‪F‬ا ال‪F‬رط‪F‬وب‪F‬ة التي تجتمع ف‪F‬يها األم‪F‬ران ج‪F‬ميعا‪ ،‬أعني الطبيعي وال‪F‬عرضي‪ ،‬ف‪F‬ال‪F‬دالئ‪F‬ل ال‪F‬دال‪F‬ة‬ ‫عليها مركبة من جنسي هذا الدالئل التي ذكرنا‪.‬‬ ‫‪ .٦‬والتح‪ّF‬رز م‪F‬ن ح‪F‬دوث ال‪F‬زك‪F‬ام ع‪F‬ن ه‪F‬ذه ال‪F‬رط‪F‬وب‪F‬ة ي‪F‬كون بس‪F‬ببني‪ ،‬أح‪F‬ده‪F‬ما أن يمنع ت‪F‬ول‪F‬ده‪F‬ا ال‪F‬بتة‬ ‫واآلخ‪F‬ر أن تج‪F‬ذب إذا ت‪F‬ول‪F‬دت في ال‪F‬رأس حتى ينقى م‪F‬نها ال‪F‬دم‪F‬اغ وي‪F‬ؤم‪F‬ن ان‪F‬صباب‪F‬ها إلى ال‪F‬صدر‬ ‫والرئة والحلق فال يحدث عنها سعال وال نزالت‪.‬‬ ‫‪ .٧‬ف‪F F‬أمنع ت‪F F‬ول‪F F‬د ال‪F F‬رط‪F F‬وب‪F F‬ات في ال‪F F‬دم‪F F‬اغ بس‪F F‬ببني‪ ،‬أح‪F F‬ده‪F F‬ما ت‪F F‬نقية ال‪F F‬بدن م‪F F‬ن ال‪F F‬فضول وزي‪F F‬ادة‬ ‫األخ‪F‬الط التي ت‪F‬رتفع إلى ال‪F‬دم‪F‬اغ م‪F‬ن س‪F‬ائ‪F‬ر ال‪F‬بدن ل‪F‬ئال ي‪F‬قبل م‪F‬ن ال‪F‬بدن ال‪F‬فضول التي يمتلى ب‪F‬ها‬ ‫>واآلخر تقوية الدماغوم‪F‬ن ال‪F‬عاق‪F‬رق‪F‬رح‪F‬ا‬ ‫وزن ‪ م‪F‬ن ق‪F‬وة< م‪F‬مسكة >و< ق‪F‬وة م‪F‬لطفة وأي‪F‬ضا ق‪F‬وة مس‪F‬ددة وق‪F‬وة م‪F‬فتحة وأي‪F‬ضا م‪F‬ن ق‪F‬وة ج‪F‬اذب‪F‬ة وق‪F‬وة‬ ‫دافعة‪.‬‬ ‫‪ .٤٩‬ف ‪F‬ما ك ‪F‬ان ت ‪F‬رك ‪F‬يبه م ‪F‬ن ه ‪F‬ذي ال ‪F‬قوتي امل ‪F‬تضادة ال يخ ‪F‬لو م ‪F‬ن أن ي ‪F‬كون ال ‪F‬قوت ‪F‬ان م ‪F‬تساوي ‪F‬تني أو‬ ‫أح‪F‬ده‪F‬ما أق‪F‬وى‪ .،‬ف‪F‬إن ك‪F‬ان ال‪F‬قوت‪F‬ان م‪F‬تساوي‪F‬تني في الشيء ال‪F‬ذي ه‪F‬و م‪F‬ما ق‪F‬وت‪F‬ان ل‪F‬ه‪ ،‬ف‪F‬ال يخ‪F‬لو م‪F‬ن‬ ‫أن تكون طبيعة ذلك الشيء مناسبة إلحدى القوتني أو مخالفة لهما جميعا‪.‬‬ ‫‪ .٥٠‬ف‪F F‬إن ك‪F F‬ان‪F F‬ت م‪F F‬ناس‪F F‬بة إلح‪F F‬دى ال‪F F‬قوت‪F F‬ني |‪ |103b‬ظه‪F F‬ر ف‪F F‬عل ت‪F F‬لك ال‪F F‬قوة على ال‪F F‬قوة االخ‪F F‬رى‬ ‫فنسب الفعل إليها‪.‬‬

‫‪283‬‬

‫‪Qusṭā b. Lūqā. On protection against rheum and catarrhs‬‬

‫‪ .٥١‬و إن ك‪F‬ان ال‪F‬قوت‪F‬ان م‪F‬ختلفتني ل‪F‬قوة ال‪F‬دواء امل‪F‬رك‪F‬ب وال ت‪F‬ناس‪F‬بان‪F‬ها وال واح‪F‬دة م‪F‬نها ظه‪F‬ر في ذل‪F‬ك‬ ‫املركب فعل متضاد كالذي يظهر من فعل الورد‪.‬‬ ‫‪ .٥٢‬ف‪F‬إن‪F‬ه م‪F‬رك‪F‬ب م‪F‬ن ق‪F‬وة مح‪F‬للة وم‪F‬نحفة وق‪F‬وة م‪F‬لززة م‪F‬لبدة‪ ،‬وق‪F‬وت‪F‬ه املح‪F‬للة تح‪F‬لل ال‪F‬بخارات وامل‪F‬واد‬ ‫ع‪F‬ن ال‪F‬دم‪F‬اغ‪ ،‬وق‪F‬وت‪F‬ه امل‪F‬لبدة وامل‪F‬لززة تمنع >م‪F‬ا< ح‪F‬للته ال‪F‬قوة املح‪F‬للة م‪F‬ن الخ‪F‬روج فيخ‪F‬رج ع‪F‬ن ذل‪F‬ك‬ ‫ال‪F F‬زك‪F F‬ام‪ ،‬ألن الشيء ال‪F F‬ذي تح‪F F‬لل وط‪F F‬لب الخ‪F F‬روج إذا وافى م‪F F‬ان‪F F‬عا ي‪F F‬منعه م‪F F‬ن الخ‪F F‬روج ف‪F F‬يراجع‪،‬‬ ‫وتصادم البخار النافذ والراجع فيحدث عن ذلك الزكام‪.‬‬ ‫‪ .٥٣‬ول‪F‬ذل‪F‬ك م‪F‬ن نقى م‪F‬ن ال‪F‬ورد أق‪F‬ماع‪F‬ه حتى يبقى ورق‪F‬ه وح‪F‬ده‪ ،‬ل‪F‬م يح‪F‬دث ع‪F‬ن اس‪F‬تنشاق‪F‬ه زك‪F‬ام ألن‬ ‫القبض أكثر في أقماعه والتحليل أكثره في ورقه‪.‬‬ ‫‪ .٥٤‬ول‪F‬ذل‪F‬ك م‪F‬ا ك‪F‬ان يس‪F‬تعمل ألم‪F‬ير امل‪F‬ؤم‪F‬نني امل‪F‬توك‪F‬ل على اهلل في ن‪F‬واف‪F‬ذه وم‪F‬جال‪F‬سه م‪F‬ن ال‪F‬ورد ورق‪F‬ه‬ ‫فقط‪ ،‬منقى من األقماع‪.‬‬ ‫‪ .٥٥‬ف‪F‬قد ج‪F‬معنا في ه‪F‬ذا ال‪F‬كالم إلى ص‪F‬فات م‪F‬ا يتح‪F‬رز ب‪F‬ه م‪F‬ن ح‪F‬دوث ال‪F‬زك‪F‬ام في الش‪F‬تاء ت‪F‬قوي‪F‬ة‬ ‫الرأس والدماغ بال عنف وال أذى‪.‬‬ ‫‪ .٥٦‬وه ‪F‬ذا املعنى ك ‪F‬ان آخ ‪F‬ر م ‪F‬ن امل ‪F‬عاني التي أم ‪F‬ر أن ن ‪F‬كتب م ‪F‬ا ي ‪F‬حتاج ف ‪F‬يه‪ ،‬وق ‪F‬د بقي م ‪F‬ن ت ‪F‬لك‬ ‫امل ‪F‬عاني معنى ب ‪F‬ال ‪F‬ثال ‪F‬ث وه ‪F‬و ب ‪F‬ال ‪F‬عالم ‪F‬ات التي تس ‪F‬تدل ب ‪F‬ها على زي ‪F‬ادة امل ‪F‬رار األس ‪F‬ود في ال ‪F‬بدن‬ ‫وطريق عالجه‪ ،‬و إّنا نبني ذلك في مقالة ثانية‪ ،‬إن شاء اهلل‪.‬‬ ‫‪ .٥٧‬ت ‪F‬م ال ‪F‬كتاب في ال ‪F‬زك ‪F‬ام وع ‪F‬لله وأس ‪F‬باب ‪F‬ه وع ‪F‬الح ‪F‬ه والتح ‪F‬رز م ‪F‬نه‪ ،‬والح ‪F‬مد هلل ك ‪F‬ثيرا وصلى اهلل‬ ‫على خير خلقه محمد وآله االكرمني وسلم تسليما كثيرا‪.‬‬

‫‪Translation‬‬ ‫‪Qusṭā b. Lūqā, the Greek, on protection against rheum and catarrhs11 that occur‬‬ ‫‪in the winter.‬‬ ‫‪In the name of God, the merciful and compassionate‬‬ ‫‪1. On the best prophylactic regimen for rheum caused by a large quantity of‬‬ ‫‪moisture in the head, what cleanses the stomach and draws [matter] from the‬‬ ‫‪head without violence, the signs that indicate that there is black bile in the body‬‬ ‫‪and in what way this is treated, what can be used to strengthen the head and‬‬ ‫‪the chest and the limit that should not be exceeded when heating, cooling,‬‬ ‫‪moistening and drying. I have recorded, for each of these themes, what is‬‬ ‫‪sufficient, God willing.‬‬ ‫‪11 For the concepts zukām (rheum, coryza) and nazalāt (catarrh, defluction), see Qusṭā, Safar al‬‬‫‪Ḥaǧǧ, Bos’ commentary nos. 149 and 150, p 112. Oribasius discusses catarrh and coryza in‬‬ ‫)‪Synopsis IV (Daremberg V pp. 728ff.‬‬

284

Lena Ambjörn

2. Concerning the regimen that protects against the occurrence of rheum due to a large quantity of moisture in the head, I say that moisture in the brain must either be natural or accidental, or both natural and accidental. 3. As for the natural moisture, the indication is that the hair on the head is lank, that there is neither baldness nor any reduction of the hair on the top of the head, that the person sleeps a lot and tolerates sleeplessness badly, that he is often affected by forgetfulness and that when he sleeps, his sleep is heavy and waking up is difficult. 4. As for the accidental moisture, the indication is increased snuffling, that tears are shed easily and quickly and that the person is prone to be affected by rheum due to causes that generate moisture, such as |99b| the winter season and a moistening regimen when it comes to food and drink. 5. As for moisture that is a combination of the two, I mean the natural and the accidental, it is indicated by a combination of the two sets of signs that we just mentioned. 6. Prophylaxis against the occurrence of rheum from this moisture is accomplished by two means: one is to prevent it from being generated in the first place, and the other is that it is drawn out, once it has been generated in the head, so that the brain is cleansed from it, and make sure that it does not flow to the chest, lungs and throat, so that it does not cause cough or catarrh. 7. Generation of moisture in the brain is prevented by two means: One is to cleanse the body from residues and superfluous humours that rise to the head from the body, so that it will not receive, from the body, residues that fill it, . 8. As for cleansing the body, it can be done in two ways: one is to restrict the food, whichever the food, so that no superfluous matter is accumulated in it, and the other is to expel extant residues by phlebotomy, intake of a purgative drug and purgation by means of moderate exercise and constant baths. 9. As for phlebotomy, if the purpose is to cleanse the whole body, one should use the widest veins, those most filled and most protruding, and the amount of blood let out should be in accordance with the age of the person, the external appearance of his body and his habits and regimen. 10. As for taking a drug to cleanse the whole body, one should use strong drugs composed from simples that cleanse the body from all superfluous humours. 11. The best compounds for this kind of cleansing is the pill that Galen composed, known as the qūqāyā, and next to it the pill known as the stomachic – the

Qusṭā b. Lūqā. On protection against rheum and catarrhs

285

greater, the medium and the lesser – and next to these pills the decoctions made from myrobalan, Belleric and Emblic myrobalan, agaric, turpeth and hiera. These drugs cleanse the body in such a way that the rising of moist humours from the body to the brain is prevented, [humours] that might flow to the chest and cause catarrh and rheum. 12. As for what strengthens the brain so that it does not accept matter that comes from superfluous food, which it is too weak to digest and which results in accumulation of residues |100a| in it, it is to balance its mixture and equalise its four natural powers: attraction, retention, digestion and excretion. 13. The mixture of the brain is balanced and its powers equal if its constitution is originally strong and equal, and no accidental injuries that disrupt the equality and balance have occurred. 14. If the mixture of the brain exceeds the natural state, either originally or because of a condition that has affected it later, it is necessary to examine the qualitative change of its natural mixture, be the change simple or complex, and counteract it with opposite drugs and regimen until it becomes balanced. Thus the accumulation of residue in it is avoided, as well as the occurrence of rheum and catarrhs due to matter flowing from it. 15. As for if residue is generated in it because it is too weak to digest all the matter that comes to it from the food, the residue should be drawn out from it and expelled from it to avoid that, with time, a large amount of water from the food accumulates in it and passes on to the respiratory organs and causes coarse catarrhs, or gets lodged in the olfactory organs and causes rheum, or gets lodged in the vocal organs and causes hoarseness or cough. 16. The drugs that cleanse the brain are of two kinds: one is drugs that are swallowed, and which attract moisture from the brain and expel it through diarrhea, and the other is drugs that are chewed or gargled with, and which attract moisture from the brain and expel it from the mouth or as secretion from the nose. 17. The drugs that are swallowed cleanse the head and attract residues from the brain, and most effective of these is whatever is composed from aloe and aromatic substances, and the best compound that works in this way is the bitter hiera. 18. Since some people detest aloe, one can use myrrh instead of aloe, for myrrh is similar to aloe when it comes to mixture and effect.

286

Lena Ambjörn

19. Yaḥyā b. Māsawayh12 has mentioned this concerning myrrh and he composed a hiera for those who have hemorrhoids but cannot tolerate |100b| taking aloe, and he used myrrh instead of aloe although myrrh is less effective than aloe, since aloe has more purgative power than myrrh. 20. The most vitalizing drugs, and those that have the strongest effect when it comes to cleansing the head from moist residues – the expulsion of which from the brain, in the autumn, protects from the occurrence of rheum and catarrhs in the winter – are all swallowed, with aloe, in the autumn, as we have prepared it, and this is the description: one takes thirty dirhams of yellow myrobalan without stones, pounded and sieved, and two raṭls of boiling water is poured over it and it is stirred and clarified. Ten dirhams of good, sweet-scented, easily crumbled, transparent aloe socotrina of the best quality is added, and nutmeg, mastic, clove, cinnamon and mace, one dirham of each. These drugs are coarsely crushed, the aloe is well pulverized, and all this is put in a bottle together with the myrobalan extract and suspended in the sun. 21. One takes, on the second day after it was suspended in the sun, one and a half ūqiya with one dirham of sweet almond oil, and on the third day two and a half ūqiya, and on the fourth three ūqiya. Then one keeps on taking a quarter of a raṭl every day for totally seven days, with the amount of almond oil that we have specified. 22. If the person detests aloe and can by no means take it, colocynth pulp can be used instead of aloe. It is prepared according to this description: One takes twenty dirhams of yellow myrobalan and pounds it coarsely, and five dirhams of yellow, light colocynth pulp, cleansed from peel and stones and neither musty nor discoloured and cuts small pieces with a scissors and mixes this with the crushed myrobalan and the aromatics that we mentioned in the preceding description, i.e. mastic, nutmeg, clove, cinnamon and mace, one dirham of each, coarsely pounded and added to the |101a| myrobalan and colocynth pulp. Three raṭls of boiling water is poured over it all and it is placed in a bottle and suspended in the sun. One takes it with sweet almond oil in the amount that we have specified in the previously mentioned description of how to prepare aloe, and in the way it is prescribed there when it comes to the number of days and the daily dosage. 23. The coarse residue can be attracted from the head and the entire body by taking myrobalan with colocynth pulp, and this is the description: One takes 12 Yuḥannah (Yaḥyā) b. Māsawayh, lat. Mesuë, 777–ca 857.

Qusṭā b. Lūqā. On protection against rheum and catarrhs

287

ten dirhams of yellow myrobalan, pulverized and sieved through silk. Two ūqiyas of boiling water is poured over it, it is stirred and clarified, and this is repeated a second time. The water is collected and one adds to it one or two dāniqs of crushed colocynth pulp, which has been carefully selected as we described above, depending on the age of the one who takes it, his strength and the mixture of his body. It is taken in the evening, lukewarm, God willing. 24. I have tried this drug many times on young boys, adolescents and old men, and I highly recommend it for cleansing the entire body, and particularly the head, the brain and the parts. 25. The colocynth drug, which we have also composed, cleanses the body and the head thoroughly, which protects against catarrhs and rheum, and the description is that one takes twenty-five dirhams of myrobalan, twenty-five dirhams of white, hollow, resinous turpeth and five dirhams of blue, shiny, delicate, easily crumbled and strongly smelling scammony. These ingredients are pounded separately, sieved through silk and mixed well in a mortar. One takes two dirhams of it and kneads it with half an ūqiya of skimmed honey, two ūqiyas of hot water are poured over it and it is dissolved and taken in the evening. 26. We have composed this drug, too, and used it on a large number of people, and we can, indeed, recommend it. It is one of the most potent drugs for cleansing the head as well as the entire body. 27. The head can be specifically cleansed with the pill called šabyār and the so called ‘gold pill’ and other pills composed of aloe and aromatics, but we have left out |101b| the descriptions of these, since that is superfluous and since there is rarely any need to use them. 28. With what we have mentioned of these compounds, and since I, for each one of them, note how it works in general, followed by the particular case, I would like to continue what was mentioned above concerning prophylaxis against rheum, saying that the regimen that I hope will protect from rheum and catarrhs in the winter should be carried out in the autumn in this way: 29. He should start at the beginning of autumn by taking myrobalan water with the hiera of Theodoretus. The quantity of the hiera should be estimated with regard to what one can expect him to tolerate. This is followed by letting out blood from vena medialis, and the extraction of blood by phlebotomy is followed by the colocynth-drug, one dose every tenth day.

288

Lena Ambjörn

30. If he adds colocynth pulp to every second dose of myrobalan water during that number of days – I mean colocynth pulp from the last day of the colocynth drug and from taking the myrobalan water – that would be good. 31. If he detests colocynth pulp, he can substitute it with one dirham of pulverised agaric in the myrobalan water. The agaric should be white, light, from the inner part of a large piece, not from the bark, and it should crumble easily and have a strong scent. One dirham of it is kneaded with skimmed honey and mixed with the myrobalan water and taken. 32. When he has taken three of four doses of these two drugs – I mean the colocynth drug and the myrobalan water – or of one of them, at the intervals of days that we have specified, he should let out blood from vena cephalica at the end of autumn, followed by taking myrobalan with [the hiera of] Theodoretus. If he uses the hiera of Archigenes or the hiera of Logadios with the myrobalan instead of the hiera of Theodoretus, that would be very good and protect more strongly against the occurrence of catarrhs and rheum in the winter, God willing. 33. This is what needs to be established concerning prophylaxis against rheum and catarrhs in the winter, and what cleanses the head by gargling and what is chewed belong here as well. 34. What can be used as treatment in order to attract the residues from the head is a gum prepared from stavesacre, pyrethrum and olibanum. The description is |102a| that one takes one dirham of stavesacre and […] of pyrethrum and six dirhams of olibanum. The stavesacre and the pyrethrum are pulverised, the olibanum is melted in the fire until it becomes soft, and the two dry drugs are kneaded into it. One takes one dirham of this and it is chewed with whatever has accumulated in the oral cavity for as long as it takes for the gum to melt. Of what is chewed, this is the best gum one can use in order to cleanse the head, the uvula, the mouth and its adjacent areas. 35. As for gargels that attract from the head, the most potent of them is this gargel: One takes mustard, two dirhams, hyssop, five dirhams, stavesacre, five dirhams, water-mint and pennyroyal, five dirhams of each, marjoram and thyme, five dirhams of each, and the bast of caper root, two dirhams, and rue, twenty dirhams. This is put together in a vessel and six raṭls of refined vinegar is poured over it. This is boiled until reduced by half, clarified, and a corresponding amount of honey is poured over what is left, and this is boiled until reduced by a quarter. One gargles with it every morning for several days.

Qusṭā b. Lūqā. On protection against rheum and catarrhs

289

36. The use of fumigants with hot water cleanses the head from superfluous phlegm and bad vapours, providing the body is cleansed and that the residues are found in the brain only. 37. The best one can use of the waters used for fumigation is the fumigant that we will describe: One takes camomile, melilot, wormwood, southernwood, rue, fig-leaves and seed of harmel, twenty dirhams of each. This is brought together in a bottle and twenty raṭls of water are poured over it. The top of the bottle is firmly closed, it is boiled profusely and the warm water is poured into a concave basin. The head is covered with a shawl and one leans over the basin, where the water is, until the vapour rises to the head and causes profuse sweaing. 38. This is repeated several times – after the body has been cleansed by phlebotomy and by the intake of a drug, for if the vapour is used on a filled body it will draw more from the head to the body than it dissolves from it, but if used after cleansing the body with drugs and phlebotomy |102b| it will potently dissolve the cold, coarse residues in the head. 39. The brain can be cleansed and the residues heated and evacuated from it, and the coarse [residues] softened, with a fumigant that should be inhaled according to this description: One takes Indian aloeswood, unripe and good, five dirhams, ambergris, one dirham, laudanum, two dirhams, fār misk, five dirhams, storax, i.e. the fumigant that is imported from the Roman empired and called ‘Roman storax’, thirty dirhams, and sandal-wood, two dirhams. The dry ingredients are crushed, each one separately, the fār misk is cut into small pieces and all of this is mixed. The storax is mixed with hot skimmed honey, the dry ingredients are kneaded into it and hazelnut-shaped pills are made from it. 40. One of these hazelnut-shaped pills is placed on smoldering embers. A funnel, turned upside down, is placed over it and the tip of the funnel is inserted into the nose so that the vapours rise through the pipe of the funnel, passing from the nose to the brain. 41. When the head has become filled by it and further fumigation through the nose is not tolerated, remove the tip of the funnel for a short while in order to rest, then reinsert the tip of the funnel into the nose. This is done several times; it strengthens a weak brain, heats a cold brain and dissolves the coarse residues and digests the soft residues accumulated in it. 42. Another fumigant can be composed which is less effective than this one when it comes to cleansing the brain but more effective when it comes to evacu-

290

Lena Ambjörn

ating the residues from it. The description is that one takes sandarac, amber, ungues odorati, olibanum and galbanum, five dirhams of each, and bitumen and asphalt, i.e. Judean bitumen, ten dirhams of each, and bdellium, twenty dirhams. This is crushed and mixed, and one fumigates with it in a small house where there is no air and the door is closed. One adds of this fumigant, for every kānūn, what increases its power little by little so that its fragrance can be inhaled and reach the brain. 43. This fumigant, besides dissolving from the head and concocting the cerebral residues, improves bad air and prevents epidemic diseases |103a| and plague, and it is on the whole the most appropriate one against harmful effects of air. 44. This kind belongs to the category of fumigants called qūfī, which are the ones by which the Greeks fumigated the planets. There are seven descriptions, one for each planet. We have not mentioned them here, since much has been written about them and since the two fumigants that we have described replace them all and cleanse better than any of them. 45. Sweet-smelling fragrances strengthen the brain and dissolve bad vapours from it, and the best to use in this regard are the perfumed cremes prepared from sweet-smelling, aromatic substances, and inhalation of dissolving fragrances such as narcissus, jasmine and marjoram. White naphtha acts like that, as does castoreum and roasted black cumin. Other drugs that have a strong scent, act gently and penetrate easily act like that, too. 46. The brain can also be cleansed and strengthened by things that cause sneezing, the most potent of these is struthion. 47. Generally, all fragrant things, cold or warm, strengthen the brain, and strongly smelling things – regardless of whether they are sweet or unpleasant, as long as the smell that is inhaled from them is hot and strong – will dissolve the coarse residues accumulated in the brain, concoct raw residues and draw out bad humours. 48. This is the place to warn against scents that actually generate rheum, such as rose, for I know several people – one of whom is the Commander of the Faithful, al-Mutawakkil ʿala Allāh – who get rheum from the smell of roses. This is because the substance of the rose and [some] other plants and fruits is a combination of different or even opposite powers: a retentive and a refining power, and also an obstructive power and an opening power, and also an attracting power and an excreting power.

Qusṭā b. Lūqā. On protection against rheum and catarrhs

291

49. In whatever is a combination of two such opposite powers, the powers must either be equal or one of them stronger. If the two powers are equal in something which has two [opposite] powers, then the nature of this thing must either agree with one of the two powers or disagree with both of them. 50. If it agrees with one of the two powers |103b| the effect of this power will conquer the other power, since the effect corresponds to [the nature]. 51. But if both powers disagree with the power of the compound drug and neither of the two agrees with it, this compound will show an opposite effect, as shown by the effect of the rose. 52. For it is a combination of a dissolving, attenuating power and an adhesive, consolidating power, and while the dissolving power dissolves vapours and matter from the brain, the adhesive, consolidating power prevents that which the dissolving power has dissolved from being let out, and this causes rheum. For if something that has been dissolved and demands to be let out is met by an obstacle that prevents it from being expelled, so that it flows back and the onflowing vapour and this reflux collide, this causes rheum. 53. However, if one removes the perianth from the rose so that only the petals remain, the smell will not induce rheum, because astringency dominates in the perianth and dissolution dominates in the petals. 54. This is why one uses, for the Commander of the Faithful, al-Mutawakkil ʿala Allāh13, in his windows and session rooms, only the petals of the rose, without the perianth. 55. In this treatise we have discussed how the occurrence of rheum in the winter is prevented by strengthening the head and the brain without violence and harm. 56. This theme was the last of the themes about which he has commissioned us to write down whatever is necessary. Of these themes, the third – on the signs by which we conclude that there is an increase of black bile in the body, and how to treat it – has been left out since we will explain that in another treatise, God willing. 57. The book on rheum, its causes and presuppositions, treatment and prophylaxis, is completed. Abundant praise to God and may he bless his most excellent creature Muḥammad and his noble people and grant them salvation.

13 r. 847–861 in Samarra.

292

Lena Ambjörn

Materia medica, Arabic–English–Latin afāwīh aromatic substances. The five principal aromatic substances are musk, ambergris, aloeswood, camphor and saffron. Māsawayh AS, pp. 6 and 9. aġārīqūn agaric, Fomes officinalis. Dtr III 1. amlaǧ Emblic myrobalan, Phyllantus emblica. Ibn Ǧulǧul Ergänz. no. 7. ʿanbar ambergris, excrement of Physeter macrocephalus. Ibn Ǧulǧul Ergänz. no. 34. ʿāqirqarḥā pyrethrum, Anacyclus pyrethrum. Dtr III 69. ʿasal honey, Mel communis. Dtr II 64. aẓfār aṭ-ṭīb snail shell opercula, Ungues odorati. Dtr II 6. bābūnaǧ chamomile. Dtr III 129. baḫūr fumigant, incense. Fellman, Qalānisī pp. 178ff. balīlaǧ Belleric myrobalan, Terminalia bellerica. Ibn Ǧulǧul Ergänz. no. 6. bārzad galbanum, resin of Ferula galbaniflua. KA no. 30 basbāsa mace, seed-case of Myristica fragrans. Dtr I 50. bunduq hazelnut; for hazelnut-shaped pills, see Fellman, Qalānisī, pp. 181f. dārṣīnī cinnamon. Dtr I 10, 11. dawāʾ ḥanẓalī colocynth drug, a compound medicament composed by Qusṭā, cf. Adwiya Mushila 5:9. See also ḥanẓal. fārat al-misk fār misk, a kind of musk. See KA no. 217, 218. fūdhanǧ ǧabalī pennyroyal, Nepeta cataria. Dtr III 35 n. 3 fūdhanǧ nahrī water-mint, Mentha aquatica. Dtr III 35 n. 2. ġarġara gargle. Fellman, Qalānisī, pp. 196ff. ǧauz buwwa nutmeg, Myristica fragrans. KA no. 69. ǧundībādastar castoreum, excretion from Castor fiber (beaver). Dtr II 22. ḥabb here: pill. Fellman, Qalānisī, pp. 202–205. ḥabb adh-dhahab “gold-pill”. Budge I 263f. ḥabb aš- šabyār aloe pill, Fellman, Qalānisī, pp. 206f., recipes in Dustūr p. 30; Qusṭā, Adwiya Mushila 10:10. halīlaǧ myrobalan, the fruit of the Terminalia chebula-tree. Ibn Ǧulǧul Ergänz. no. 1. ḫall vinegar. Dtr I 69 n. 16. ḫamr wine ḥanẓal colocynth, Citrullus colocynthis. Dtr IV 167. ḫardal mustard, Brassica nigra. Dtr II 138. ḥarmal harmel, wild rue, Peganum harmala. Dtr III 46. ḥumar asphalt, bitumen. Maim. No. 168. iklīl al-malik melilot. Dtr III 40. ʿilk mastic; gum. Ḥāwī 21 no. 555.

Qusṭā b. Lūqā. On protection against rheum and catarrhs

293

iṣṭūmāḫīqūn stomachic, Stomachikon, a compound purgative pill. Fellman, Qalānisī, p. 207. iṣṭurak dry storax, Styrax officinalis. Dtr I 28. iyāraǧ from Gr. hiera, “sacred”. A compound purgative with many ingredients, in the form of an electuary. Fellman, Qalānisī, pp. 214–217. Qusṭa, Adwiya Mushila, 5:5, mentions eleven kinds of hiera. iyāraǧ arḫāǧānis the hiera of Archigenes. iyāraǧ fiqarāʾ hiera picra, bitter hiera. iyāraǧ luġādhiyā the hiera of Logadius. iyāraǧ tāʾudūrītūs the hiera of Theodoretus. kā(h)rubā amber. Dtr I 49 n. 4. kabar caper, Capparis spinosa. Dtr II 156. kundur olibanum, dried resin from Boswellia Carteri. Dtr I 30. kundus struthion, a sternutative. Dtr II 147. lādhan ladanum, resinous juice of Cistus. Dtr I 66. laḫlaḫa a perfumed composition. Fellman, Qalānisī, pp. 2221f. lauz almond, Prunus amygdalus. Dtr I 99, 100. maiʿa liquid storax. Māsawayh AS pp. 23f. maiwīzaǧ stavesacre, Delphinium staphisagria. Dtr IV 141. marzanǧūš marjoram, Origanum majorana. Dtr III 39. maṣṭakā mastic, resin from Pistacia lentiscus. Dtr I 34. maṭbūḫ decoction. Fellmann, Qalānisī, pp. 245f. misk musk, gland secretion from Moschus moschiferus. Ibn Ǧulǧul Ergänz. no. 33. mūmiyāʾī pitch, bitumen, Mumia naturalis. Dtr I 39. muql bdellium, an aromatic resin from Commiphora africana et al. Dtr I 29. murr myrrh, a resin from Commiphora abyssinica. Dtr I 27. nafṭ abyaḍ white naftha. nammām thyme, Thymus serpyllum. Dtr III 38. narǧis narcissus, Narcissus poeticus. Dtr IV 147. qafr al-yahūd Judean bitumen (= ḥumar). Dtr I 38. qaranful clove, Carophyllus aromaticus. Dtr III 44 n. 7. qayṣūm southernwood, Artemisia abrotanum. Dtr III 25. qūfī a compound fumigant of egyptian origin. Fellmann, Qalānisī, pp. 205f. qūqāyā “pill”, a compound purgative attributed to Galen. Fellmann, Qalānisī, pp. 205f. Recipe in Qusṭā Ḫadar, ch. 6. ṣabr aloe, sap from the leaves of e.g. Aloë arborescens and Aloë ferox. Dtr III 23. šabyār: see ḥabb

294

Lena Ambjörn

sadhāb rue, Ruta graveolens. Dtr III 45. šaḥm al-ḥanẓal colocynth pulp, see ḥanẓal. ṣandal sandal-wood, Santalum. Dtr I 41. sandarūs sandarac, resin of Thuya orientalis or articulata. Dtr I 21. saqmūniyā scammony, Convolvulus scammonia. Scammonium is dried juice from the root: “Der aus der dicken fleischigen Wurzel gewonnene eingedickte Milchsaft, das Scammonium, bildet unregelmässige, zerbrechlige, rothbraune oder schwärzliche, auf dem Bruche glänzende, in Splittern durchscheinende Massen.” Berendes p. 464; Dtr IV 160. šīḥ wormwood, Artemisia. Dtr III 26. šūnīz nigella, fennel flower, black cumin. See Dtr III 74. tīn fig, Ficus carica. Dtr I 107. turbid turpeth, Ipomoea turpethum. Dtr IV 121. ʿūd hindī Indian aloeswood. Kahl, Dispensatory. ward rose, Rosa. Dtr I 68. yāsmīn jasmine, Jasminum. Dtr III 40 n. 7. zūfā hyssop, Hyssopus off. Dtr III 27.

Materia medica, English–Arabic agaric aġārīqūn almond lauz aloe ṣabr aloe pill ḥabb aš-šabyār amber kā(h)rubā ambergris ʿanbar aromatic substances afāwīh asphalt ḥumar bdellium muql bitumen mūmiyāʾī; ḥumar caper kabar castoreum ǧundībādastar chamomile bābūnaǧ cinnamon dārṣīnī clove qaranful colocynth ḥanẓal colocynth drug dawāʾ ḥanẓalī colocynth pulp šaḥm al-ḥanẓal

Qusṭā b. Lūqā. On protection against rheum and catarrhs cumin, black šūnīz decoction maṭbūḫ fig tīn fumigant baḫūr galbanum bārzad gargle ġarġara gold-pill ḥabb adh-dhahab gum ʿilk harmel ḥarmal hazelnut bunduq hiera iyāraǧ hiera, bitter iyāraǧ fiqarāʾ hiera of Archigenes iyāraǧ arḫāǧānis hiera of Logadius iyāraǧ luġādhiyā hiera of Theodoretus iyāraǧ tāʾudūrītūs honey ʿasal hyssop zūfā Indian aloeswood ʿūd hindī jasmine yāsmīn Judean bitumen qafr al-yahūd ladanum lādhan mace basbāsa marjoram marzanǧūš mastic maṣṭakā; ʿilk melilot iklīl al-malik musk misk mustard ḫardal myrobalan halīlaǧ myrobalan, Belleric balīlaǧ myrobalan, Emblic amlaǧ myrrh murr naftha, white nafṭ abyaḍ narcissus narǧis nigella šūnīz nutmeg ǧauz buwwa olibanum kundur pennyroyal fūdhanǧ ǧabalī

295

296

Lena Ambjörn

perfumed composition laḫlaḫa pill ḥabb pitch mūmiyāʾī pyrethrum ʿāqirqarḥā rose ward rue sadhāb rue, wild ḥarmal sandal-wood ṣandal sandarac sandarūs scammony saqmūniyā snail shell opercula aẓfār aṭ-ṭīb southernwood qayṣūm stavesacre maiwīzaǧ stomachic [pill] iṣṭūmāḫīqūn storax, dry iṣṭurak storax, liquid maiʿa struthion kundus thyme nammām turpeth turbid vinegar ḫall water-mint fūdhanǧ nahrī wine ḫamr wormwood šīḥ

Bibliography Berendes = Des Pedanios Dioskurides aus Anazarbos Arzneimittellehre in fünf Büchern. Übersetzt und mit Erklärungen versehen von Julius Berendes. Stuttgart 1902. Budge = Syrian anatomy, pathology, and therapeutics; or, “The Book of Medicines”, the Syriac text; edited from a rare manuscript with an English translation, etc. by Ernest A. Wallis Budge. Oxford 1913. Correspondance = Une Correspondance islamo-chrétienne entre Ibn al-Munaǧǧim, Ḥunayn ibn Isḥāq et Qusṭā ibn Lūqā, edited, translated and commented by Paul Nwyia and Khalil Samir, in: Patrologia Orientalis 40/4, _Turnhout 1981. Dtr = Dioscurides Triumphans. Ein anonymer arabischer Kommentar (Ende 12. Jahrh. n. Chr.) zur Materia medica. Arabischer Text nebst kommentierter deutscher Übersetzung herausgegeben von Albert Dietrich. Göttingen 1988.

Qusṭā b. Lūqā. On protection against rheum and catarrhs

297

Dustūr = ad-Dustūr al-Bimāristānī. Le Formulaire des Hôpiteaux d’Ibn Abil Bayan, ed. Paul Sbath, Bulletin de l’Institut d’Égypte, Cairo 1933. Gabrieli, Giuseppe, Nota biobibliographica su Qusṭā ibn Lūqā, in: Rendiconti della Reale Accademia dei Lincei, Classe die scienze morali, storiche e filologiche, Serie V, vol. 21 (1912) 341–382. GAS = Fuat Sezgin, Geschichte des arabischen Schrifttums, Bd I–IX. Leiden 1967ff. Gutas, Dimitri, Greek Thought, Arabic Culture. London 1998. Fellmann, Qalānisī = Irene Fellmann, Das Aqrābādhīn al-Qalānisī. Quellenkritische und Begriffsanalytische Untersuchungen zur arabisch-pharmazeutischen Literatur. Beirut 1986. Hinz, Walther, Islamische Maße und Gewichte. Umgerechnet ins metrische System. Leiden 1955. Ḥunayn, Mā turǧima = Ḥunain ibn Isḥāq über die syrischen und arabischen Galen-Übersetzungen, ed. Gotthelf Bergsträsser. Leipzig 1925. Ibn Ǧulǧul, Ergänz. = Sulaimān ibn Ḥassān al-Andalusī, Ibn Ǧulǧul, Die Ergänzung Ibn Ǧulǧuls zur Materia medica des Dioskurides, Arabischer Text nebst kommentierter deutscher Übersetzung herausgegeben von Albert Dietrich. Göttingen 1993. Ibn Ǧulǧul, Ṭabaqāt = Sulaimān ibn Ḥassān al-Andalusī, Ibn Ǧulǧul, Ṭabaqāt al-aṭibbāʾ wal-Ḥukamāʾ (Les générations des médecins et des sages), ed. Fuʾād Sayyid. Cairo 1955. Ibn an-Nadīm, K. al-Fihrist, ed. G. Flügel, Bd I–II. Leipzig 1871–72. Ibn al-Qifṭī, Taʾrīḫ al-Ḥukamāʾ, ed. Julius Lippert. Leipzig 1903. IAU = Ibn Abī Uṣaibiʿa, K. ʿUyūn al-anbāʾ fī ṭabaqāt al-aṭibbāʾ, 2 vols., ed. August Müller. Kairo-Königsberg 1882–84. KA = Yaʿqūb b. Isḥāq al-Kindī, The medical formulary or Aqrābādhīn of al-Kindī, tr. by Martin Levey, Madison 1966. Kahl, Dispensatory = The Dispensatory of Ibn at-Tilmīd. Arabic text, English translation, study and glossaries by Oliver Kahl. Leiden 2007. Maimonides = Šarḥ asmāʾ al-ʿUqqār (L’Explication des noms de drogues) Un Glossaire de Matière médicale compose par Maïmonide. Texte publié pour la première fois d’après le manuscript unique avec traduction, commentaires et index par Max Meyerhof. Cairo 1940. Māsawayh AS = Traité sur les substances simples aromatiques par Yohanna ben Massawaih, ed. Paul Sbath. Bulletin de l’Institut d’Égypte 19 (1937) 5–7. Oribasius = Œuvres d’Oribase, ed. Bussemaker et Daremberg, 6 vols., Paris 1851–73. Qusṭā, Adwiya Mushila = Qusṭā ibn Lūqā on Purgative Drugs and Purgation. Kitāb Qusṭā ibn Lūqā fi l-Adwiya al-Mushila wa-l-ʿilāǧ bi-l-Ishāl. Edition, translation and commentary by Lena Ambjörn. Frankfurt am Main 2004. Qusṭā, Ḫadar = Qusṭā ibn Lūqā on Numbness. A Book on Numbness, its Kinds, Causes and Treatment according to the opinion of Galen and Hippocrates. Edition, translation and commentary by Lena Ambjörn. Stockholm 2000. Qusṭā, Safar al-Ḥaǧǧ = Risāla fī tadbīr safar al-ḥaǧǧ. Qusṭā ibn Lūqā’s medical regime for the pilgrims to Mecca. Edition, translation and commentary by Gerrit Bos. Leiden 1992.

298

Lena Ambjörn

Savage-Smith, Emilie, “Some Sources and Procedures for editing a Medieval Arabic Surgical Tract” in: History of Science 14 (1976) 245–264. Siegel, Rudolf, Galen’s System of Physiology and Medicine. An Analysis of his Doctrines and Observations on Bloodflow, Respiration, Humours and Internal Diseases. Basel 1968. Temkin, Owsei, Galenism: Rise and Decline of a Medical Philosophy. Ithaca 1973. Wilcox, Judith C., The Transmission and Influence of Qusṭā ibn Lūqā’s “On the Difference between the Spirit and the Soul”. New York 1985.

Aesthetics of generosity – generous aesthetics On the cultural encoding of an Arab ‘national virtue’ Stephan Guth, University of Oslo

Introduction During his long stay in Paris in the 1820s, the head of the Egyptian study mission, Rifāʕa Rāfiʕ al-Ṭahṭāwī (1804–1873), observes, among many other things, that the French “show charity (musāwāt) only in words and deeds, not when it involves their money and possessions. While they do not refuse to lend something to their friends when asked, they never give things away, except if they are certain of obtaining some form of recompense. In truth, they are avaricious rather than generous.” The reason for that, al-Ṭahṭāwī continues in his travelogue, is the simple fact that “generosity is peculiar to the Arabs”1 – as, by the way, outlined already earlier in a chapter on hospitality in a translation of his.2 This quotation makes clear that karam, noble generosity, was looked upon by an Arab scholar of the first half of the 19th century as a kind of Arab national virtue long before it was to become, with increasing Western penetration of the Middle East, an important ingredient in the cultural comparison between East and West: Western materialism (māddiyyat al-ġarb) vs. the true, ‘spiritual’ values of the East (rūḥiyyat al-šarq).3 And even though East-West confrontation has

1  

Ṭahṭāwī [1993] II: 149 / tr. Newman 2011: 179.  

2

According to Newman 2011: 179, fn. 2 (and 144, fn. 2), this Muḫtaṣar al-siyar wa’lʕawāʔid (Abridgement of the Conducts and Customs) is an abridged version of alṬahṭāwī’s translation of George-Bernard Depping’s Aperçu historique sur les mœurs et coutumes des nations (Paris 1826), made earlier under the title Qalāʔid al-mafāḫir fī ġarīb al-ʔawāʔil wa’l-ʔawāḫir (in Newman’s rendering: ‘Exquisite Poetry of the Glorious Qualities of the Strange Customs of Those that Come First and Those that Come Last’; translation completed Nov. 1829, published as book in Būlāq 1833).

3

For a detailed account of this discussion in modern Arabic fiction, see Wielandt 1980: esp. 372-382 (mostly on Tawfīq al-Ḥakīm’s ʕUṣfūr min al-šarq), but see also Index s.v. “Materialismus des Westens im Gegensatz zum Spiritualismus des Ostens, as well as El-Enany 2006 (see entries “materialism”, “matter-spirit dichotomy”, and “spiritualism” in the Index); for the māddiyya–rūḥ(ān)iyya divide as a general idea

300

Stephan Guth

been leveled now, in a postcolonial age of globalization, generosity and hospitality are still readily quoted, alongside with intact family structures and respect for the elderly, as positive ‘markers’ of Middle Eastern cultures, whenever the adjustment or correction of a one-sided negative image of Islam is on the agenda. But how could al-Ṭahṭāwī, whose writings still remained rather unbiased by the polarization that is so typical of colonialism, an author who wrote long before the advent of Arab nationalism and the creation of the modern nationstates, already have the idea of generosity as a specifically Arab national virtue? In the following I will try to demonstrate that al-Ṭahṭāwī’s attitude can be explained as the result of an age-old, mainly linguistic, and especially literary, aestheticization of generosity that can be traced back as far as to pre-Islamic times, an aestheticization that is hardly conceivable without the assumption that the social acts of generosity themselves were in some way experienced ‘aesthetically’. I was inspired to look into this connection between generosity and aesthetics by Aida Kanafani’s and Annegret Nippa’s short articles on “Rites of Hospitality and Aesthetics” (1983/1993) and “Art and Generosity” (2006), respectively, from which it became clear to me that the ‘cultural institution’ of generosity with all probability cannot be fully understood only from the four angles from which it is usually approached – as a set of social acts, as a means to negotiate social order, as expressing emotions (love, respect, etc.), and as a complex sphere of ethics. Rather, there is a fifth component involved – beauty. Thus, anthropologist Kanafani holds that “aesthetic aspects of hospitality [are evaluated] much as we might evaluate a play or a musical performance,”4 and art historian Nippa who, in her observation of hospitality rituals among Bedouins, noticed a “primacy of the procedure over consumption,” concluded that “[h]ospitality underwrites the main field of aesthetic activity among the Bedouins” and that in Arab generosity, beauty “supports the representation of honor, self-consciousness and the family.”5 I will illustrate this idea, first, with a look into pre-Islamic poetry, then into a few anecdotes about the generous hero of Arab tradition par excellence, the pre-Islamic poet Ḥātim al-Ṭāʔī, as these produced a larger corpus of texts within the so-called ʔadab literature. in modern Arab thought, cf. Gershoni / Jankowski 1995/2002: esp. ch. 2, “‘Now is the turn of the East’: Egyptian Easternism in the 1930s” (pp. 35ff.). 4

Kanafani 1993: 128 (my italics – S.G.).

5

Nippa 2006: 557 and 564 (my italics – S.G.).

Aesthetics of generosity – generous aesthetics

301

A. Pre-Islamic Poetry Generosity cannot rightly be said to figure as an Arab ‘national virtue’ in preIslamic poetry – there are only generous individuals. However, generosity emerges clearly from this corpus of texts as one of the most important elements of Bedouin culture and ethics:6 it is, among other things, through munificence and large-handedness that a clan chief demonstrates his manly virtue (murūʔa7), his chivalry,8 high standing or ‘nobility’9 (the old term for this was ḥurriyya10), his ability to care for the poor and needy and to offer protection and safety to those traveling outside their tribal areas (the institution of hospitality was of the highest importance for trade).11 Generosity also could be taken as an indication of a clan chief’s ability to ward off the risks posed, in principle, by any stranger since excessive hospitality served the purpose of calming poten-tially hostile strangers by putting them in a favourable, benevolent mood. A similar function can still be observed today. In her study on hospitality rituals of women in 6 According to Bonner (2003: 21), “[t]he best discussion so far of generosity in preIslamic Arabia has been M[eïr] M. Bravman’s essay ‘The Surplus of Property’ [i.e., Bravman 1962].” 7 On the term cf., e.g., Farès 2012, with further references. – See also Izutsu 1966: 75, where murūʔa / muruwwa, which “included such various virtues as generosity, bravery and courage, patience, trustworthiness, and truthfulness,” is described as “the highest ethical ideal of the Jāhilīyah.” 8 In Izutsu’s words: “[generosity] was first and foremost an act of chivalry. A man who could make a royal display of his generosity was a true dandy of the desert. Generosity in this sense was a master passion of the Arabs. It was not so much a ‘virtue’ as a blind, irresistible impulse that was deeply rooted in the Arab heart” – Izutsu 1966: 76. 9 No wonder then that the notions of generosity and nobility are hardly separable – “the adjective karīm is just the word in Old Arabic for such a combination of the ideas of lavish generosity and nobility” – Izutsu 1966: 76. 10 Cf. a verse attributed to the legendary embodiment of Arab generosity, the preIslamic poet Ḥātim al-Ṭāʔī: (sarīʕ) ʕasà yarà nāraka man yamurrū / ʔin ǧalabat ḍayfan fa-ʔanta ḥurrū “perhaps a passer-by sees your fire / If it brings a guest, you are ḥurr” – reported, among others, by Zamaḫšarī [n.d.]: 26. – The term ḥurriyya came to be used to denote the modern concept of ‘freedom’ when the ideas of the French Revolution were received in the Arab world, cf. Monteil 1960 and Rebhan 1986 (see indices), and particularly Rosenthal / Lewis 2012. 11 As Bravman 1962 has shown, “[t]he Arabs thought and said that property (mâl) has a surplus (faḍl, or ʕafw), which its owner must give away. Indeed, the surplus carries a right or claim (ḥaqq) within itself” – summarized by Bonner (2003: 21).

302

Stephan Guth

modern Golf countries, Kanafani speaks of the need to “neutralize” the power of the guest by certain behavioral standards.12 In a Bedouin context, Annegret Nippa called the famous traditional coffee ceremony an “act of pacification.”13 (Cf. also the fact that the Arabic word for “guest”, ḍayf, perhaps goes back to an Afroasiatic root *ĉ̣ayVp- which with some probability signified both guest and stranger.14 A similar ambiguity15 is well-known also from Greek xénos, Latin hostis, and the French adjective étrange.16) By means of generosity the Bedouin also secured his own good reputation and in this way legitimized the power he exercised in the tribal society. In poetry, we encounter a generosity that is already aestheticized: it is both a subject of poetry and attired in, and vested with, a highly elaborate, ornate language, i.e., it has been transformed into a product of aesthetic culture, elevated onto a poetic, and that is at the same time: a quasi-magical, level.17 This demon12 Kanafani [1983] 1993: 131. 13 Nippa 2006: 550. – For an earlier description of this ceremony, and the “Rules of Hospitality” in a rural setting of the early twentieth century in general, cf. elBarghûthî 1924. – As for the rites of hospitality as “acts of pacification”, cf. also Chelhod’s interesting observation that Arab poets “ne disent presque rien de l’hôte, celui-ci n’ayant, à leurs yeux, qu’un rôle passif et secondaire. Celui qui reçoit l’hospitalité est à la fois un émir, un prisonnier et un poète, disent les Bédouins. Il est d’abord un prince, car il a droit à tous les égards; mais en échange, il ne doit se conduire en obligé. Il est interessant de noter que le verbe adâfa [i.e., ʔaḍāfa], donner l’hospitalité, signifie aussi ajouter, annexer, lier. Celui qui est reçu est prisonnier de qui le reçoit, car il doit se conformer en tout aux directives du maître du lieu” – Chelhod 1990: 13. – Very similar in a Persian context: “De fait, l’invité perd son indépendance en devenant, comme dit ironiquement un dicton persan, ‘la bourrique du maître de maison’ [the host’s jenny]!” – Richard 1990: 33. 14 Orel & Stolbova (1994: #584) reconstruct the Afroasiatic form as the common ancestor of Semitic *ṣ̂ayp- ‘guest’ (reconstruction based on Arabic ḍayf, and also ḍayf in Ḥarsūsī and Mehrī; cf. also Jibbālī eḍef ‘to give hospitality’), West Chadic *ĉ̣ay(V)p- ‘friend; pilgrim, stranger; guest’ (based on forms like mi-zɛp, mә-zɛp, mɛzɛp, mɛ-dap, sâpa, mi-zı̂va, n- zafe in several languages) and Central Chadic *mi-sı̂p‘guest’ (reconstructed from actual forms like misı̂pi, mәsә̂bi, misı̂bi, mihibi). 15 It goes without saying that it is an ambiguity only from our modern perspective and for the modern perception – for the pre-Islamic Arabs, Greeks, etc. it was the mutual relationship that counted and, because it involved the host and the guest likewise, was expressed with one and the same word. 16 As noticed, among others, by Derrida 1997. 17 As mentioned, among others, by Wagner 1987 I: 33, an aesthetic appreciation of poetry as a form of art should be seen, originally, in connection with, or based on,

Aesthetics of generosity – generous aesthetics

303

strates how close generosity and aesthetics, i.e., social action and art/literature, are interrelated in this culture and that generosity to a large extent participates in this ‘magic’. The fixed, almost unalterable, quasi-eternal forms (monorhyme, monometre, etc.) and the standardized ideal images into which real/factual acts of generosity are translated in this poetry, reflect the correlation, in pre-Islamic life, of (formalized, institutionalized) generosity, a firm social order, ethical normativity, and the ‘magic’ and beauty of the ritualistic, ‘formulaic’ language: as strict and regular as the poetic forms, but also as enchanting as the heroically idealizing verses, are also the codes of giving, the rituals of generosity, and their social functions.18 The high importance of generosity in pre-Islamic Bedouin society – according to the poet Ṭarafa it has the capacity to disguise and hide any vice one may think of19 – correlates with the sheer amount of verses on generosity that have come onto us, and the richness, diversity and sophistication of the correspondits magical function (“... wird die künstlerische Wertschätzung der Poesie wohl nie völlig von der magischen Funktion zu trennen gewesen sein”). 18 Cf. Hamori who (1974: 22) remarks that “[q]aṣīda poets spoke in affirmation of a model they shared; their poetry tended to become a shared experience, all the more as the affirmation was through the replay of prototypal events which the model so successfully charted.” Therefore, “[t]he organization of the material within the qaṣīda too ha[d] a ritualistic aspect”. It is for this reason that also generosity was expressed “formulaically” (ibid.: 23). 19 Ṭarafa (as quoted by Nanah 1987: 47): wa-yuẓhiru ʕayba ’l-marʔi fī ’l-nāsi buḫluhū * wa-yasturuhū ʕanhum ǧamīʕan saḫāʔuhū // taġaṭṭa bi-ʔasbābi ’l-saḫāʔi fa-ʔinnanī * ʔarà kulla ʕaybin wa’l-saḫāʔu ġiṭāʔuhū “Das Geizen läßt die Schande unter den Leuten sichtbar werden und die Freigebigkeit bedeckt sie vor allen Leuten. Bedecke dich mit den Mitteln der Freigebigkeit, denn ich finde, daß jede Schande die Freigebigkeit als Hülle hat” (what discloses a man’s meanness against / disgrace among the people is his stinginess, while liberality conceals it completely from them. Cover yourself with the means of liberality, for I think that any disgrace finds its cover in liberality). Similarly also Zuhayr b. ʔAbī Sulmā, Muʕallaqa, v. 50 (in other countings: 42, or 51), in Septem Moallakat, ed. Aug. Arnold, Leipzig 1850 (as quoted by Izutsu 1966: 76): “Whoever makes of generosity a shield for his personal honor makes it grow. But whoever neglects to guard himself from blame, will be blamed” (Arabic: wa-man yaǧʕal-i ’l-maʕrūfa min dūni ʕirḍihī / yafirhu, wa-man lā yattaqi ’lšatma yuštamī; another translator renders the verse as “And he, who makes benevolent acts intervene before honor, increases his honor; and he, who does not avoid abuse, will be abused” – Charles F. Horne, ed., The Sacred Books and Early Literature of the East, vol. 5: Ancient Arabia: The Hanged Poems; The Koran, New York 1917: 39).

304

Stephan Guth

ing vocabulary. Suffice it to throw a quick glance only at the variety of verbs that all have to do with giving openhandedly: baḏala, ǧāda, rafada, ḥabā, saḫā, samaḥa, ʔaʕṭà, ʔaʕāna, fāda, manaḥa, ʔanāla, ʔahdà, wahaba;20 cf. also, in addition to the corresponding verbal nouns (esp. ǧūd, saḫāʔ, samāḥa), the nouns karam and qiran (det. qirà).21 Cf. also the many words that are known for the competing with a rival in slaughtering camels (and other precious animals), a ritual that reminded already Stetkevych22 of the so-called potlatch rituals of North American west coast Indians where whole tribes may be ruined by their leaders’ excessively giving away own property in order to outrival a competitor in social (and political) rank: muʕāqara / taʕāqur (to cut the tendons), munāḥara (to cut the throats), mufāḫara / tafāḫur, mumāǧada / miǧād (to compete for glory and fame), taʕāṭī (to compete in giving).23 Apart from the persons involved in acts of generosity and the special circumstances under which the heroes display and give proof of their generosity, it is first and foremost the inventory of the gifts that is described with poetic opulence: the camels and noble horses; the (female) singers and slaves; deliciously smelling, grease-dripping fried camel humps; the enormous amounts of wine or milk; the glowing, visible from far away, of the welcoming fire lit on a hill; the cauldron; the eating bowls, always filled with the best pieces of meat; the drinking vessels that are steadily refilled; the huge heaps of ashes that testify to the largeness of the fire that had been lit and, hence, to the extent of the host’s generosity, and so on.

20 Nanah 1987: 277ff. 21 Interestingly, qiran, as well as the corresponding verb ʔaqrà, seem to be etymologically related to qarya, pl. quran (det. qurà) “village” (Orel / Stolbova 1994: #1568 give Ugaritic qr-t, qry-t, Hebrew qiryā, Syriac qerī-t-, South Arabian qr, Jibbālī sirɛ-t as cognates within Semitic and reconstruct Sem. *ḳary- ‘town, village’). – For the etymology and semantic history of generosity-related vocabulary cf. my study on this issue, . 22 So Bonner 2003: 29, note 37, with reference to J. Stetkevych, Muhammad and the Golden Bough: Reconstructing Arabian Myth, Bloomington: Indiana UP, 1996. 23 Cf. Nanah 1987: 134 f., Bonner 2003: 19–21.

Aesthetics of generosity – generous aesthetics

305

Add to this verbal opulence the images, expressed in similes, metaphors or metonymies, that serve to highlight, on the one hand, the quantity of what is given away openhandedly (it is often a ‘sea’, an ‘inundation’, a ‘flash flood’, a ‘torrent’ that is pouring down on the beneficiary of generosity).24 On the other hand, besides quantity also the frequency of giving is emphasized, and the quality of the gift is equated with the elemental life-giving power of water in the desert: most frequently, the effect of generosity is compared to that of rivers, brooks, rain, or the spring.25 The word nadà ‘humidity, dew’ is so popular that it becomes virtually synonymous with ‘generosity’,26 and so on. In connection with the large-handed distribution of own property verbs like ‘to ruin’, ‘to destroy’, ‘to annihilate’ (ʔatlafa, ʔafnà, ʔahlaka, etc.) are also frequently met,27 used with extreme proud, of course – an indication, again, of the similarity, mentioned above, between pre-Islamic generosity rituals and American Indian potlatch ceremonies and their often devastating effect: not seldom, whole tribes could be run into the ground.28

24 For examples and attestations cf. Nanah 1987: 208ff. 25 Ibid. 26 Cf. entry nadan (det. nadà) in Freytag 1837 IV: 262 (√ndw): “[...] Pluvia; Humor, mador, humiditas [...] ; Pabulum [...] ; Liberalitas ; [...].” – For an example, chosen rather randomly, cf. a verse bei al-ʔAʕšā: ḏākum-u ’l-māǧidu ’l-ǧawādu ʔAbū ’lʔAšʕaṯi ʔahlu ’l-nadà wa-ʔahlu ’l-suyūbi “dies ist der ruhmvolle, gütige ʔAbû-l-ʔašʕaṯ, der Herr der Freigebigkeit und der Herr der Gaben” – al-ʔAʕšā / Geyer 1905: 139. 27 Cf. Nanah 1987: 292ff. 28 A reflex of this may be Qurʔānic wa-ʔanfiqū fī sabīli ’llāhi, wa-lā tulqū bi-ʔaydī-kum ʔilā ’l-tahlukati “And spend in the way of Allah, and do not cast yourselves to destruction with your own hands” (Q 2: 195, tr. Maulana Muhammad Ali, ed. Zahid Aziz, Ahmadiyya Anjuman Lahore Publications, 2010: 45). In this verse, the second imperative (lā tulqū...) can be read as a warning to everybody who wants to make a charitable donation not to exceed the limits and thereby bring ruin over oneself. For other possible readings see however Paret 1980-II: 40-41, who states that “[e]s ist nicht klar, was der (in sich eindeutige) Satz in diesem Zusammenhang bedeuten soll”. He mentions that, according to al-Ṭabarī, the ancient exegetes allowed for four different interpretations and that Fischer (1911) connected it to the ǧihād, which is the general topic of the surrounding verses. Paret himself, however, seems to favour the version he adds (and which is close to the reading mentioned in the beginning of this footnote): “Man könnte aber auch übersetzen bzw. paraphrasieren: ‘Aber stürzt euch nicht ins Verderben (indem ihr euch durch übermäßig hohe Spenden verausgabt)!’” – Paret 1980 II: 41.

306

Stephan Guth

The attitude, observable in pre-Islamic poetry, towards generosity and the high esteem in which it was held, remains more or less unaltered also in the period following Islamization. Kaʕb b. Zuhayr, for instance, praises in his burda the prophet Muḥammad and the Qurayš for their bravery, generosity and noble descent in the same way as also pre-Islamic poets would have done.29 From here, the ideal of generosity is transmitted farther into the Umayyad period and reappears there, despite the emergence of new poetic genres, as a popular topic, as for example in hunting poetry, one of the most booming genres of the time, in scenes where the successful hunter, inviting to dinner, demonstrates how liberally and generously he gives away to the guests. The fundamental continuity remains unaffected also by the fact that the potlatch-like excessive slaughtering competitions now become outlawed (they will continue to be practised for quite a while, though30). Furthermore, the new religion, Islam, is eager to adopt, if not usurp, the positively connoted semantic field of generosity and in this way does not only contribute to a perpetuation of the ideal, affirming it as a high value per se; it also modifies this ideal, bracketing the ‘show-off’ aspects and dazzling quantity by instead underlining the finer, milder, more charitable components:31 genero29 Cf. Izutsu 1966: 75: “It is quite significant [...] that, in the pictures of Muḥammad which the pious Muslim writers of later ages have left, we often see a typical hero of the Arabian desert. Interestingly enough, the personal characteristics attributed to Muḥammad in the books of Tradition are quite in line with the old nomadic ideals of man that we find so highly praised in the works of pre-Islāmic poets. Take for example the following description of the personality of the Prophet by ʕAlī b. Abī Ṭālib, given by Ibn Hishām in the Sīrah: ‘He was of all men the most openhanded [!], most stout-hearted, most truthful of tongue, most loyal in the keeping of his trusts, most serene of mind, and the noblest in friendly intercourse. Those who saw him for the first time feared him, but those who got acquainted with him loved him. Indeed, a man like him I have never seen.’ This is nothing but a picture of an ideal man, containing no element at all that might have been repugnant to the moral sense of a Jāhilī Arab.” 30 Two hundred camels are reported to have been slaughtered, as late as under the reign of the forth caliph, ʕAlī b. ʔAbī Ṭālib (r. 656–661), by the father of the poet alFarazdaq (c. 641–c.730). Their consumption, however, was forbidden by the caliph since the slaughtering had been “consecrated for the worship of something other than God” – Bonner 2003: 21. 31 Cf. Izutsu 1966: 77–78: “Islām denied all value to acts of generosity originating in the desire to make a show. Dandyism or chivalry for its own sake was in this view nothing but a satanic passion. [...] In order that generosity may become a genuine

Aesthetics of generosity – generous aesthetics

307

sity and liberality are no longer privileges of the pre-Islamic Bedouins; now, it is, above all, God himself who is ǧawād and karīm, and also the Qor’an is a qurʔān karīm.32 In addition, the duty, inherent in the pre-Islamic ethical code of generosity, to provide for the poor and needy, is now transferred from the individual leader or clan chief to the community of believers as a whole: although alms given voluntarily and generously by individuals (ṣadaqa) are still honorable and meritorious, they should at any rate correspond reasonably to the individual’s property and are altogether subordinate to zakāt, the general alms tax that is one of the ‘Five Pillars’ of Islam, a main duty incumbent on every Muslim.33 As is well known, the expansion, soon after its ‘birth’, of Islam into territories far beyond its ‘cradle’ brings along intense contact with many other civilizations and a strong mutual influencing between conquerors and conquered, a contact and influencing that are again more intense than the encounters and exchanges that had been common already long before the rise of the new religion and, in their totality, had formed what is now generally referred to as the culture of Late Antiquity. In the course of the conquests, the conquered civilizations, on the one hand, undergo a process of Islamization and Arabization. Islamic virtue, it must first of all be deprived of the thoughtlessness which characterized it in the days of Jāhilīyah. One who goes to the length of slaughtering on the spur of the moment, or worse still merely for display, all his camels without stopping a moment to think that his act may reduce him and his family to misery and perdition on the morrow – such a one may very well have been a model of murūwah or karam in Jāhilīyah, but is no longer to be considered a man of true generosity. A man of true generosity is he who ‘expends his wealth in God’s way’, that is, from a pious motive. And being founded on piety, it must be something well-controlled and restrained. Generosity in Islām is something essentially different from the boastful and excessive charity of which the pagan Arabs were so fond.” 32 For “The Islamization of Old Arab Virtues” in general, comprising also courage, loyalty, veracity, and patience, cf. Izutsu 1966: 74–104 (ch. 5). 33 For the etymology and semantic history of ṣadaqa and zakāt cf. Weir 1994 and Zysow 2001, respectively, as well as my own study, mentioned above (fn. 21). – On the integration, and ‘reform’, of pre-Islamic virtues in Islam cf. Izutsu 1966: 75: “What is [...] important to note [...] is that Islām did not revive or restore these nomadic virtues as it found them among the Bedouin. In adopting and assimilating them into its system of moral teachings, Islām purified and freshen-ed them, making their energy flow into certain channels which it had prepared” (my italics, S.G.). On zakāt, see ibid.: 78ff.

308

Stephan Guth

Islam, on the other hand, adopts and integrates innumerable techniques and concepts from these civilizations into its own system of knowledge and values – the huge number of loanwords in the Arabic language are the best witness to that. At least equally important as the techniques and concepts, however, is the fact that Islam is also able to integrate the staff themselves, the specialists and experts in the various fields. And it is, first and foremost, the group of Persian secretaries, the kuttāb, in most cases associated as ‘clients’ (mawālī) to Arab masters, to whom we owe that ‘refinement’, realized through recourse to Persian high court culture, of old Arabic customs and ethical codes, a refinement that led to the (re-)formation of the concept of ʔadab that was to become one of the key concepts – if not the most important at all – of classical Arabic and Islamic civilization.34 34 It is equally correct to speak of a ‘formation’ and a ‘re-formation’: considering that fact that there was a pre-Islamic ʔadab we are certainly dealing with a re-formation of an older concept; the many new elements in the re-formed concept that come in addition to its old characteristics, however, allow us to speak, with the same right, of post-conquest ʔadab as if it was a new concept. – The question of the cultural origins, and hence also the etymology, of ʔadab is still unsolved. While Arab lexicologists are eager to derive the term from ʔadaba, impf. yaʔdubu ‘to invite (to a repast, a banquet, i.e. a maʔduba)’ “because it invites men to the acquisition of praiseworthy qualities and dispositions, and forbids them from acquiring such as are evil” (as Lane I 1863: 35 summarizes the Classical positions in English), a common theory in Western scholarship is that the word is a secondary formation from the pl. ʔādāb, which is not from a sg. ʔadab, but from daʔb ‘a custom’, and hence originally was *ʔadʔāb (first put forward by Nallino, cf. Gabrieli 1960/2008). Ilse Lichtenstädter (1974, quoted by Horst 1987: 208) thought it could be an idea to trace it back to a Sumerian é-dub-ba-a, signifying “school” or “university”, while Asbaghi 1988 – probably motivated by Iranian national pride – proposed a confluence, in Arabic (and Persian) ʔadab, of two Middle Persian (abb. mPers) ancestors: the first, meaning “gute Sitten, Anstand, Höflichkeit” (good manners, consideration, politeness), is traced back to mPers aīvēn (which in turn, the author says, goes back to an Old Iranian *abi-dagna); the second, meaning “literarische Bildung” (literary formation, knowledge of/from literature), has its root, according to Asbaghi, in mPers dipi “Inschrift” (inscription) (from Old Iranian *dipi-vara). Rolland 2014, s.v., gives Sumerian DUB > Akkadian ṭuppu ‘tablet, sacred text’ or Old Persian dipi ‘to write’ as the most plausible alternatives. In contrast, other recent research tends to look at the region where the Arab term emerged, as a multicultural space, the Middle East of Late Antiquity (cf. Neuwirth 2010), where it is difficult, or impossible, to identify one single origin and where one rather has to assume the influence of several ‘national’ traditions, among which also the key concepts of Persian frahang and ēwēn/ āyīn as well as the Greek paideía. On frahang and ēwēn/āyīn, cf., briefly, Khaleghi-

Aesthetics of generosity – generous aesthetics

309

B. Prose anecdotes The re-formation of the old concept of ʔadab, now often translated as “refinement”, or “culture”, “civilization”, or “humanitas” in general, and in particular its ‘codification’ in written form, i.e., its literarization, did not remain without consequences for the future of generosity discourses and, hence, also the concept as a whole. One of the most important processes in this regard is the embedding of transmitted verses, among which also those with a generosity topic, into prose narratives, short “news” (ʔaḫbār, sg. ḫabar, where historicity, the historical probability of the reported events is emphasized), in anecdotes (nawādir, lit. “rariMotlagh 1983 who says that “Adab is the equivalent of the Middle Persian frahang and New Persian farhang (T. Nöldeke, “Geschichte des Artaschir-i Pāpakān aus dem Pahlevi übersetzt mit Erläuterungen und einer Einleitung versehen,” Bazzenberger’s Beiträge zur Kunde der indogermanischen Sprache 4, 1879, p. 38, note 3; H. S. Nyberg, Hilfsbuch des Pehlevi, Uppsala, 1931, II, p. 70); it is also very close to another Pahlavi word, ēwēn, Persian āyīn, meaning custom, rule, correct manner, and the like. Thus in certain Arabic works of the early Islamic centuries, ēwēn is rendered either by adab and its pl. ādāb, or by rasm and its pl. rosūm; but sometimes the original word, in its Persian form āyīn, is retained.” – For paideía, the standard reference is Jaeger 1933–1947. The main traits of the conceptual history of this term are summarized also in The New Pauly: paideía “is the main Greek term for the education [...] of a child (paîs, παῖς) and above all of a young person [...] The prerequisite for paideía that goes beyond unconscious sozialisation [sic!] is the concept that if phýsis (‘talent’) exists, the aretḗ (‘being good’) can be acquired by means of the contemplation and practical imitation of models, for which the téchnai (‘arts and handicrafts’) provide the model. Paideía which consists of gymnastic and musical elements, serves to mould the child according to the (originally aristocratic) ideal of kalokagathía (‘external and internal excellence’).” The concept received its ‘classical’ shape through Isocrates [436–338 BC], whose paideía programme that “geared towards the cultivation of the human capacity for communication and at the same time – as it was believed – towards ethical moulding, [was] also useful in its humanistic orientation as a resource for the individualistically shaped ideal of personal development in the Hellenistic period. [...] To the Greeks of the Hellenistic period, paideía [...] was considered to be ‘the most precious good that is given to mortals’ [...]; in it they see their cultural identity defined [...]. The Romans appropriated the term paideía as humanitas [...]. The ideas of paideía survived through their appropriation by the Romans, and it is this form that can be regarded as the first Humanism” – Christes 2006. A good idea of the similarity of concepts in the ‘globalized’ Middle East of Late Antiquity can be gained, for example, from the entry on “Gastfreundschaft” (hospitality) in Reallexikon für Antike und Christentum, see Hiltbrunner / Gorce / Wehr 1972.

310

Stephan Guth

ties”, sg. nādira, where the focus lies on the extraordinary, the particularly noteworthy), or in simple “stories” (ḥikāyāt, sg. ḥikāya).35 In this way, the pre-Islamic verses are not only preserved and conserved for posterity, they are also provided with situational contexts. Taken together, ʔaḫbār, nawādir, ḥikāyāt etc. form a large amount of usually quite short, exemplary narratives about individual persons acting in very specific, concrete situations, a corpus of ʔadab stories that corresponds to, and displays a high degree of similarity with, its religious counterpart, the sunna of the Prophet as transmitted in individual ʔaḥādīṯ (sg. ḥadīṯ, “Prophetic tradition, Hadith, narrative relating deeds and utterances of the Prophet and his Companions”36), assembling the collective experience of past generations for the use by contemporaries – a casuistic ethics, so to speak, from which orientation for one’s own present can, and should, be derived. Through their ‘adabization’,37 ‘adabification’,38 or ‘adab-tation’,39 stories about munificent persons are further transmitted and ‘immortalized’, and generosity is in this way again reinforced as something with an ‘extratemporal’, ‘eternal’ aesthetic quality, an essentially beautiful cardinal virtue that is not only worth to be talked and rhymed and written about but also to be encyclopediacized. The many individual traditions in ʔadab encyclopedias, whole book chapters and even some monographs on generosity (e.g., al-Tanūḫī’s al-Mustaǧād min fiʕalāt al-ʔaǧwād) or its counter-concept, avarice or niggardliness (cf. alǦāḥiẓ’s famous K. al-Buḫalāʔ), bear eloquent witness to this. Another effect of the adabization of generosity narratives is their dissemination beyond the narrow Bedouin environment. Although Bedouins like Ḥātim al-Ṭāʔī remain prominent, they now receive company of other generous protagonists, in particular rulers and patrons. Accordingly, there is a lot of variation now in the settings, or ‘theatres’, where generosity ‘happens’ and/or is staged. Coming from an urban palace environment and having been formed along the ideals of Persian high court culture, most authors of ʔadab works, themselves often Iranians, look down at the pre-Islamic Arab Bedouins and their cultural traditions with the contempt of the civilized towards the barbarians. 35 For a short characterization of the “countless short narrative texts [that are] to be found in compilations of the most varying nature,” cf. Leder / Kilpatrick 1992: 10ff. 36 Wehr / Cowan 1979, root ḥ-d-ṯ . 37 Hamori 2006: 175. 38 Sharlet 2011: 65. 39 Guth [2015].

Aesthetics of generosity – generous aesthetics

311

Therefore they treat a number of aspects of the old Arabian ethics of generosity, on the one hand, with a good deal of skepticism, criticizing and/or ridiculing in particular the self-destructive side (e.g., the excessive slaughtering competitions, mentioned above, that drove many into ‘bankruptcy’) and occasionally displaying some practices of Bedouin generosity as, in their opinion, mere calculation, essentially dishonest showmanship in which the giver does nothing but flexing his muscles.40 On the other hand, there is no doubt that Iranians also were incredibly impressed, if not intrigued by figures like Ḥātim al-Ṭāʔī – otherwise it would be hard to explain why of all countries Iran could become home to a whole heroic epic built around the Arab poet.41 From here, it also spread further into India,42 as witnessed by, inter alia, several children’s books, Ḥātim Ṭāʔī Bollywood productions and TV series:

40 In the words of ʕAbbās Bayyūmī ʕAǧlān (al-Hiǧāʔ al-ǧāhilī, Kairo 1982: 103, quoted apud Nanah 1987: 33, fn. 20), a critical source mocked the Bedouin practice saying: ʔin qarà ʔaḥaduhum ḍayfan ʕaddahā makrumatan, wa-ʔin ʔaṭʕama ʔakalatan ʕaddahā ġanīmatan, tanṭuqu bi-ḏālika ʔašʕāruhum, wa-taftaḫiru bi-ḏālika ʔašʕāruhum “if one of them [the Arabs] hosts a guest he considers this a noble deed, and when he has fed [s.o.] some food he considers this a successful foray. Their poems bear witness to this, and their verses boast it.” 41 See Forbes 1880: vi: “The Adventures of Hatim Taï have long obtained the highest popularity in those regions of Asia where the language of Persia is spoken or studied. Among those who speak the language of the original, the work is read with admiration and listened to with delight.” Forbes translated from a manuscript that “was procured in the East in 1824. It gives no notice as to the place and time of writing; but from the beauty of penmanship and general accuracy of expression, I [i.e., Forbes] am inclined to think that it was executed in Persia [...]; and from its appearance, I should consider it at least a century old” (ibid., p. vii). 42 ... and also in Europe! For reflexes in Giovanni Boccaccio’s Decamerone (mid-14th c.), cf. Thouvenin 1933.

312

Stephan Guth

Film Haatim Tai, dir. Babu Bhai Mistry (1990)43

Children’s book, Kissa Hatimtai, by Gopal Sharma, Manoj Publications (2005)44

TV series, first broadcasted on Star Plus in 200345

43 http://www.kiks.cc/2014/02/haatim-tai-1990-dvdrip-x264-ac3-esubs.html (downloaded 30 June, 2014). 44 http://www.exoticindiaart.com/book/details/kissa-hatimtai-saga-of-incredibleyoung-man-who-could-do-just-anything-to-make-others-happyIDH533/(downloaded 30 June, 2014). 45 http://www.moduscreative.co.uk/item/the-adventure-of-hatim (downloaded 30 June, 2014).

Aesthetics of generosity – generous aesthetics

313

The stories on the generosity of Ḥātim al-Ṭāʔī that I have looked at in the ʔadab works46 display structures that are quite similar to those Fedwa Malti-Douglas discovered for the miser narratives of al-Ǧāḥiẓ and al-Ḫaṭīb al-Baġdādī.47 I have found three, or four, main basic plots, all of which actually only underline what had been important already in poetry: Pattern 1: The simple act of generous giving CALL ISSUED

A call for an act of generosity is issued, Ḥātim should/must give something

=>

CALL ANSWERED

The call is answered, Ḥātim (over)fulfils his duties as a host or giver / exceeds the expectations set in him

This pattern consists of nothing more than the nucleus of a ‘generosity event’, namely an account of it. The two major ‘functions’, in a Propp’ian sense,48 of this generosity ‘syntagm’ could be described as ‘Call for generosity issued’ and ‘Call for generosity answered’. The act of giving is reported almost as if it was a routine. An example: One day, Ḥātim al-Ṭāʔī sits down and invites all those present in the camp (al-ḥilla) to drink (wine) with him, upon which approximately two hundred people come and not only are feasted but also, upon leaving, given three camels each as an additional farewell present on top of all.49 It seems that it is, first and foremost, the surprising quantity and/or the exceptional quality of the given object that are constitutive of this plot type. In it, the ideal of generosity is hyperbolically glorified and thereby confirmed and reinforced.

46 For a list of most of these, cf. van Arendonk 2011. 47 Malti-Douglas 1985. 48 Propp 1928/1968. 49 Ḥātim al-Ṭāʔī 1963): 24-25, Wormhoudt 1984: 31 (“Ḥātim ʕalā ’l-šarāb”); Šayḫū 1890: 117. – Cf. also Wormhoudt 1984: 27 (“Ḥātim wa-rakb Banī ʔAsad”), Ḥātim alṬāʔī (1963): 21–22.

314

Stephan Guth

Pattern 2: Giving against all odds a)

HURDLE

=>

CALL ISSUED

=> b)

CALL ISSUED

=>

CALL ANSWERED

HURDLE

a) A difficult / emergency situation is described, in which a call for an act of generosity is issued or, b) a call for an act of generosity is issued but Ḥātim meets serious obstacles that might prevent compliance with the call

The call for an act of generosity is answered in spite of all obstacles, Ḥātim (over)fulfils his duties as a host or giver / exceeds the expectations set in him

In this plot type, Ḥātim al-Ṭāʔī is asked to give something in a situation in which the act of giving seems to be completely unthinkable or is only hard to imagine. The ‘Hurdle’ that has to be dealt with comes as a third Propp’ian function, in addition to the ‘Call for generosity issued’ and ‘Call answered’ of Pattern 1. The most unsurmountable hurdle is, of course, given when Ḥātim is already lying dead in his grave the moment the call to be entertained as a guest is issued to him.50 None of these obstacles however can prevent him from eventually spending liberally. In the case just referred to, for instance, it is Ḥātim’s ghost who rises from the grave, steps in for the dead Ḥātim and assumes his duties as a host. – Another famous story reports how Ḥātim regally entertains the guests of his ex-wife Māwiyya despite the fact that she has left him for another man.51 A typical emergency situation is also a general famine – which, of course, does not prevent Ḥātim from sacrificing even the last animal he’s got, his favorite horse, in order to supply something to eat for a woman and her starving children.52 50 This episode is found, among other places, in The Thousand and One Nights – cf., e.g., tr. Burton 1885–1888 IV: 94–96, tr. Weil 1839 II: 800 (with Ḥātim’s ghost replaced with a vision, appearing to his son in a dream, of Ḥātim ordering him to send an animal to the caravan who had asked for hospitality at his father’s grave; in the Nights, the head of the group is not, as in most ʔadab versions, a certain ʔAbū ’l-Ḫaybarī, but a Ḥimyarite king called “Zu ’l-Kura’a” / “Dsul Kelaa”). 51 Ḥātim al-Ṭāʔī (1963): 19–21, Caussin de Perceval 1847–1848 II: 624–627, Šayḫū 1890: 111–112, Schultheß 1897: #li, Rescher 1925: 67, Wormhoudt 1984: 24ff. 52 Ḥātim al-Ṭāʔī (1963): 22–23, Caussin de Perceval 1847–1848 II: 622ff., Rescher 1925: 65, Tanūḫī [n.d.]: 28, Wormhoudt 1984: 28.

Aesthetics of generosity – generous aesthetics

315

I tend to interpret this pattern, and the emergency situations staged in it, as the articulation of a collective fear from the loss of face (honour) and power that the inability to comply with the norms and standards of generosity may entail.53 With this function, the Giving against all odds plot type is not unlike the third type, that is Pattern 3: Generosity tested/proven CALL = TEST

=>

CALL ANSWERED = TEST PASSED

In this type of stories, the focus is not on external circumstances that look as if they could be adverse enough to prevent the generous one from giving openhandedly; rather, the focus of interest here lies on people who, for one reason or another, want to put the hero (who is already known for his liberality) and his openhandedness to the test. Here, the ‘Call for generosity issued’ function goes along with, or is identical to, a ‘Test’, and the ‘Call answered’ function becomes one with passing the test (‘Test passed’ function). In one of these stories, for instance, the Emperor of Byzantium wants to know whether Ḥātim is generous to such a degree that he would be ready to bequeath his favorite horse to him. The qayṣar sends his emissary, and without the latter revealing his identity nor expressing his master’s wish, Ḥātim entertains the guest with this very same horse – since all camels happened to be far away at pasture, he had to slaughter it for the guest. When the emissary eventually discloses who he is, who he sent him and what his mission actually aims at, the horse is already eaten, and the guest has to confess that, indeed, nobody is more generous than Ḥātim: he has given his horse to the emissary even before the latter had asked for it.54 – In another story, a woman (Ḥātim’s future wife, Māwiyya bt. ʕAfzar), courted by three men, among whom also Ḥātim, disguises herself as a beggar and

53 Cf. Chelhod 1990: 15: “Avec le cérémonial et le faste, indissociables de l’hospitalité telle que l’Arabe la conçoit, celui-ci conserva toujours le souci de la ‘blancheur de sa face’, qu’il craindrait de perdre en ne réservant pas à son hôte son plus chaleureux acceuil.” 54 Ḥātim al-Ṭāʔī (1963): 24, Caussin de Perceval 1847–1848 II: 624, Šayḫū 1890: 116– 117, Rescher 1925: 64–65, Wormhoudt 1984: 30.

316

Stephan Guth

calls on each competitor, one after the other, asking for food, etc.; while the other two do not care about the poor, Ḥātim, of course, does.55 In the Generosity tested/proven plot, the hero’s excellent (and often already legendary) reputation, based on his liberality and munificence, is taken for granted, while those who put him to the test, represent a threat and possible danger to his fame. In case the one asked to provide hospitality, protection, redemption, or release (of slaves, prisoners, etc.), or alms, is unable to fulfill the expectations put in him, he has to reckon with damage to his reputation and loss of power through libel and slander and/or invective poetry (hiǧāʔ). Even to the present day, it is a common belief that nothing is worse than to become the topic of evil gossip on the tongues of the gossip agents,56 and even today Bedouins explain the efforts they make for guests, by pointing to the fact that guests, as soon as they move on, become šuʕarāʔ, “poets” who carry both good and bad reputation, fame and shame out into the world. In the words of a Jordanian Bedouin: “The host must fear the guest. As long as he sits there [and shares the food with you], he is a guest [and thus manageable / under control]; [but] as soon as he rises, he will be a poet.”57

55 Caussin de Perceval 1847–1848 II: 613–616, Šayḫū 1890: 108-09, Schulthess 1897: #1, Rescher 1925: 66–67, Wormhoudt 1984: 20ff. 56 Kanafani [1983] 1993: 133. 57 lāzim al-muʕazzib yiḫāf min aḍ-ḍayf. luma yiǧlis howa ḍayf. luma yigum howa šāʕir “The host must fear the guest. When he sits [and shares your food], he is company, when he stands [and leaves your house], he is a poet” – Shryock 2004: 36 (quoted also apud Nippa 2006: 562). – Of another kind of fear, that of not being able to fulfil the duties of a generous host, talks the Jāhilī poet Ṭarafa when he declares, boastingly, “I am not a man who lurks about fearfully among the hills. / I am here to help, whenever people call for my charity” – Ṭarafa, Muʕallaqa, v. 45, in Septem Moallakat ... (see above, fn. 18), as quoted apud Izutsu 1966: 77. – Cf. also Ḥātim’s own words (reported by ʔAbū Tammām in his Kitāb al-Ḥamāsa): ʔiḏā mā ṣanaʕti [Ḥātim is talking to his wife, Māwiyya b. ʕAfzar] ’l-zāda fa-’ltamisī lahū / ʔakīlan, faʔinnī lastu ʔākilahū waḥdī // ʔaḫan ṭāriqan, ʔaw ǧāra baytin, fa-ʔinnanī / ʔaḫāfu maḏammāti ’l-ʔaḥādīṯi min baʕdī, which Rückert translates as: “Hast du die Kost bereitet, so hol nur auch herein / den Gast, der mit mir eße, denn nicht eß’ ich allein: // Sei es ein Nachtanklopfer, sei es ein Hausnachbar; denn üble Rede fürcht’ ich nach meinem Tod fürwahr” – Rückert 2003: 284 (6. Buch, #724).

Aesthetics of generosity – generous aesthetics

317

Pattern 4 may count as a plot type in its own right, but is actually only a Combination of plottypes 2 and 3. a)

HURDLE

=>

CALL = TEST => CALL ANSWERED = TEST PASSED

b)

CALL = TEST

=>

HURDLE

a) A difficult / emergency situation is described, in which a call for an act of generosity is issued which at the same time is a test, or b) a call for an act of generosity is issued through which Ḥātim shall be tested, but he meets serious obstacles that might prevent his compliance with the call and, hence, his passing the test

However, Ḥātim manages to answer the call for an act of generosity in spite of all obstacles, (over)fulfils his duties as a host or giver / exceeds the expectations set in him, and thereby also passes the test

Example: The call for generous giving is issued to the dead Ḥātim at his grave by a certain ʔAbū ’l-Ḫaybarī (= emergency situation + test),58 or the test, mentioned above, is carried out upon Ḥātim by the Byzantine Emperor (will Ḥātim give away even his favourite horse?) while at the same time there is an emergency situation.59 It should be mentioned, incidentally, that very similar plot types are to be found also in the realm of orally transmitted so-called ‘folk literature’, or ‘folklore’. Here, too, the high esteem in which generosity is held in general, as well as the interest in the types of challenges such a high ethical imperative poses and may demand solutions for, have been the driving forces behind an uninterrupted carrying on of traditions about generosity and a steady accumulation of a thesaurus of related of stories.60

58 al-Iṣbahānī (1905) xvi: 97–98; al-Ǧāḥiẓ (n.d., Maḥāsin): 38 f.; Šayḫū 1890: 104, 113– 114; Rescher 1925: 67–68. 59 Šayḫū 1890: 116–117; Rescher (1984): 71–72. 60 Cf. El-Shamy 1995 II (Alphabetical index of Motifs): 204, where the entries on “generosity” and “generous” give such topics as “[generosity] towards enemy”, “[g.] rewarded”, “contest in [g.]”, “envious man won over by g. of his intended victim”, “test of g.”, “test of son’s g.: duty, self-interest, or true philanthropy?”, “test of true intention behind a seemingly g[enerous] act”, “which was most g[enerous]

318

Stephan Guth

As a consequence of the tradition-bound nature of ʔadab – it is, after all, the non-religious counterpart of the sunna – accounts of ethically relevant/interesting acts that a generation inherits from an earlier one are usually passed on to the next generation in a conservative manner. In this way the theme of generosity, like any other ʔadab theme, produces a branch of traditions in its own right, a fact for which again the figure of Ḥātim al-Ṭāʔī can serve as a good example. The description of Ḥātim in Ibn Qutayba’s K. al-Šiʕr wa’l-šuʕarāʔ (second half of the 9th century CE) as a “generous one and a poet of good poetry,” who “[always] was victorious and, whenever he fought, triumphed”, a person who, “when he went plundring carried away [much]; when you asked him [for something], would give [in abundance]; when he threw the arrows [i.e., played maysir], won [lit., was first]”, and a man who, “whenever he captured [someone] released [him shortly afterwards]” – kāna ǧawādan šāʕiran ǧayyida ’l-šiʕri, [...], wa-kāna ẓafiran ʔiḏā qātala ġalab, waʔiḏā ġanima ʔanhab, wa-ʔiḏā suʔila wahab, wa-ʔiḏā ḍaraba bi’l-qidāḥi sabaq, waʔiḏā ʔasara ʔaṭlaq61 – this description can be found in almost identical form, for example, in Ø al-Maydānī’s collection of proverbs (K. al-Amṯāl) from the 12th century: here it has, apparently, already clotted into a kind of lexical entry: kāna ǧawādan šuǧāʕan šāʕiran muẓaffaran, ʔiḏā qātala ġalab, wa-ʔiḏā ġanima nahab, wa-ʔiḏā suʔila wahab, wa-ʔiḏā ḍaraba bi’l-qidāḥi sabaq, wa-ʔiḏā ʔasara ʔaṭlaq “Hatim was liberal, brave, wise, and victorious: when he fought, he conquered; if he plundered, he carried off; when he was asked, he gave; when he shot his arrow, he hit the mark; and whomsoever he took captive, he liberated.”62

– husband, robber or lover?”, “which was the better son – (g[enerous]) or dutiful?”. – Cf. also ibid., entries on “Gift” (205–206), “Hospitality” and “Host” (239). 61 Ibn Qutayba 1969: 83. 62 al-Maydānī 1867: s.v. “ʔaǧwadu min Ḥātim”; the translation follows Forbes 1880: ix (although his rendering of ʔiḏā ḍaraba bi’l-qidāḥi sabaq as “when he shot his arrow, he hit the mark” is wrong – the Arabic expression relates to throwing the arrows in the maysir game). – Cf. also al-ʕAskarī 1964: 236ff. (#517).

Aesthetics of generosity – generous aesthetics

319

Ø In this form, the information is transmitted again and reappears in numerous later ʔadab anthologies and encyclopedias, e.g., late 17th century Ḫizānat al-ʔadab by ʕAbd al-Qādir al-Baġdādī: kāna ǧawādan yušbihu ǧūduhū šiʕrahū wa-yuṣaddiqu qawlahū fiʕluhū 63 – with the result that, again more than three centuries later, the entry on Ḥātim in the Arabic Wikipedia, after briefly characterising the hero as a “preIslamic poet and generous warrior, whose generosity has become proverbial,”64 continues by elaborating: kāna ǧawādan yušbihu šiʕrahū ǧūduh, wayuṣaddiqu qawlahū fiʕluh, [...] muẓaffar, ʔiḏā qātala ġalab, wa-ʔiḏā ġanima ʔanhab, [...]”, etc., – which, except for a small addition, is virtually identical to the wording in the compilations mentioned above.65   Now, having followed the development of generosity discourses from preIslamic times through ʔadab compilations up to the premodern period, the question remains to be asked how generosity became encoded as a specifically – and almost exclusively – Arab ‘national virtue’, a characteristic trait that, today, not seldom is propagated, together with hospitality, not only as a brand mark for commercial purposes in tourism, but also as a feature of pride among Arabs when it comes to self-definition in comparison with other cultures. For a first identification of ‘generosity’ and ‘Arab’, the fact that generosity was perceived by the conquered peoples as an extremely important element in the culture of the Arab-Islamic invaders may have been one of the major reasons. In Arabic poetry, which the non-Arabs came to know as a major marker of the culture of their new masters, generosity had a prominent position and, in its poetic reflexion, was even more heroicised and idealised than it was already in reality. Andras Hamori, for example, speaks of the “hieratic quality of the qaṣīda”66 in which “a segment of the total organization of experience” is acted out in a “ritualistic performance” “according to the heroic model” – what is praised is endowed with ideal qualities and, hence, aestheticized.67 63 al-Baġdādī, Ḫizānat al-ʔadab, III: 128, 12–13, quoted apud Nanah 1987: 33; with, on one occasion, slightly altered word order, also in Šayḫū 1890: 99. 64 šāʕir ǧāhilī, fāris ǧawād yuḍrab al-maṯal bi-ǧūdih. 65 Entry “Ḥātim al-Ṭāʔī” in (as of 27 Dec 2011). 66 Hamori 1974: 21–22. 67 Cf. Hamori 1974: 23–24. – Cf. also Wagner 1987 I: 179, who in his turn mentions this tendency to sublimate reality and reshape it, in poetry, according to idealised models (“Tendenz, nicht lebendige Figuren vorzuführen, sondern sie auf ein Ideal

320

Stephan Guth

The meeting of civilizations and cultures during the Islamic expansion – an encounter that brought along forms of contact that were much more intense than those which, of course, had existed previously – then created a favorable climate for the formation of certain ethnical stereotypes. The fact that large-handedness, munificence, bounteousness etc. were important ingredients also in the concepts of hospitality that were to be found with the conquered peoples,68 on the one hand facilitated the transfer and ‘introduction’ of the Arab variety of generosity; on the other hand, when conflicts between the new masters and the indigenous population arose, the differences in the conceptualization of this virtue served to delineate the ‘own’ from the ‘other’. Thus, for instance, already in al-Ǧāḥiẓ’s times the identification of ‘excessive liberality’ and ‘Arab’ and, respectively, the lack of this type of generosity among other peoples – e.g., the Byzantines who, allegedly, did not even have a word for it – seem to have been a ‘given’, a ‘fact’ of common knowledge.69 AlǦāḥiẓ himself, it is true, did not propagate a direct equation ‘Arabs = liberal’ / ‘non-Arabs = stingy’ but, rather, called for what Fedwa Malti-Douglas termed

zu reduzieren”). He adds that E. García Gómez interpreted this tendency, labelled as the “falseness” of Arabic poetry, as indication of an essential “dishonesty” (in Wagner’s translation: “Unehrlichkeit der arabischen Dichtung”). But Wagner relativizes this dictum by explaining that the ‘corruption’ of reality in poetic representation did not originate in the poet’s creativity or imagination, but was due to the normativity of the poetical medium and the genre rules that had to be obeyed even when poetic fiction did not conform to the facts (“Diese Entfernung von der Wirklichkeit entsprang jedoch nicht der schöpferischen Tätigkeit des Dichters oder seiner Phantasie, sondern im Gegenteil: dem Zwang zur Darstellung mittels vorgegebener Stereotypen, die auch dann angewandt werden mußten, wenn sie mit der Wirklichkeit nicht übereinstimmten”, ibid.). 68 “To the ancient Greeks, hospitality was a divine right. The host was expected to make sure the needs of his guests were seen to. The ancient Greek term xenia, or theoxenia when a god was involved, expressed this ritualized guest-friendship relation. In Greek society a person’s ability to abide the laws to hospitality determined nobility and social standing.” (en.wiki) 69 al-Ǧāḥiẓ, quoting a certain Ṭāhir al-ʔAsīr who held that “what points to that the Byzantines are the most stingy nation of all is the fact that you do not even find a word for munificence (ǧūd) in their language” (mimmā yudillu ʕalā ʔanna ’l-Rūma ʔabḫalu ’l-ʔumami ʔannaka lā taǧidu li’l-ǧūdi fī luġatihim isman): Malti-Douglas 1985: 156.

Aesthetics of generosity – generous aesthetics

321

an “Islamic universalism”70; on the other hand, discussing, as he did, all the widespread ethnic clichés of his time he also participated in, and contributed to, the formation of a written tradition on the subject that both spread and, in a way, also ‘cemented’ the clichés: the written medium made them more endurable. The consolidation of the clichés continued also in spite of the so-called šuʕūbiyya movement.71 As a movement that tried to counter what was experienced as arrogance from the side of the Arab ‘upper class’ and opposed the “extreme partisans of Arabism”72 who were to be found in the early Abbasid period (as a response to increased non-Arab influence), the šuʕūbiyya aimed to do justice to all ethnical groups in its ‘assessment’ of the latter and therefore argued in favour of an “equality [of treatment] (taswiya)” of non-Arabs and Arabs.73 The learned disputes of which we know, however, show that the movement’s plea for equal treatment was unable to prevent ethnical clichés from spreading further and further. Thus, writing in the second half of the 10th century, ʔAbū Ḥayyān al-Tawḥīdī (923-1023), for example, reports that the virtues and vices of certain “nations” (ʔumam) were a current topic of the ‘table talks’ of his time and that it was common to hear that “The Persians excelled in [the art of] governing a state, the Byzantines in wisdom, the Indians in magic [... ,] the Arabs, finally, in munificence, loyalty and eloquence.”74 70 According to al-Ǧāḥiẓ, every people has its own virtues and vices and it is this fact that makes a “nation” equal to others. Islam is seen as an integrative system in which each “nation” can find its niche since all “nations” are equal in the fact that they have their specificities. Malti-Douglas 1985: 157. 71 šuʕūbiyya refers to a (predominantly) literary movement in non-Arab environments of the Muslim World, esp. Iran and Andalusia, during the 8th and 9th centuries CE. It aimed to provide a response to what the non-Arabs experienced as arrogance of the conquerors towards the conquered. The point of departure in any šuʕūbiyya argument – and hence also the name for the movement’s name – was sūra 49, v. 13: yā ʔayyuhā ’l-nāsu! ʔinnā ḫalaqnākum min ḏakarin wa-ʔunṯā wa-ǧaʕalnākum šuʕūban [!] wa-qabāʔila li-taʕārafū. ʔinna ʔakramakum ʕinda ’llāhi ʔatqākum [...] “Lo! We have created you male and female, and have made you nations and tribes that ye may know one another. Lo! the noblest of you, in the sight of Allah, is the best in conduct. [...]” (trans. J. M Pickthall). 72 Nagel 1981: 51 (original German: “[die] extremen Parteigänger des Arabertums”). 73 Ibid., quoting from Ibn ʕAbd Rabbih, al-ʕIqd al-farīd, ed. ʔAḥmad ʔAmīn [et al.], 7 vols., Cairo 1940ff., IV: 403 (my translation, S.G.). 74 Ibid., quoting from ʔAbū Ḥayyān al-Tawḥīdī, K. al-ʔImtāʕ wa’l-muʔānasa, 3 parts, ed. ʔAḥmad ʔAmīn and ʔA. al-Zayn, Cairo 1939 I: 75ff. (my translation, S.G.). – Cf. a similar ‘national’ colouring in the “Pseudo-Ǧāḥiẓ” (K. al-Maḥāsin wa’l-masāwī), in

322

Stephan Guth

Stories/narratives are one of the most important forms of cultural expression in which Arabs encoded and transmitted the huge thesaurus of life experience and their ethics. Narratives of generosity, liberality, munificence etc. are part of this wealth of experience and this ethics. As evidenced in the statement, quoted in the beginning of this article, by R. R. al-Ṭahṭāwī , the idea that true generosity is “only to be found with the Arabs” must somehow have solidified in the postclassical period and been maintained over the centuries; otherwise al-Ṭahṭāwī would not have been able to mention it as a matter of course, as a fact that does not need further explanation. But al-Ṭahṭāwī is, as mentioned above, by no means the final point of the historical process: there are similar statements also from later periods and, ultimately, also from today, among which, not surprisingly, touristic advertising should be mentioned. A nexus between ‘generosity’ on the one hand and the ‘nation’ on the other is to be observed also in the Egyptian short stories analyzed recently by Bill Granara;75 here, the link is made via the symbolism of the river Nile: the Nile as a large-handed giver and, as the country’s lifeline, the symbol of Egypt. There is however also a third element in the game: revenge. But this has to be discussed in another article ...

Bibliography ʔAbū ’l-Faraǧ al-ʔIṣbahānī → al-ʔIṣbahānī, ʔAbū ’l-Faraǧ.

ʔAbū Tammām Ḥabīb b. ʔAws → Rückert, Friedrich.

ʔAlf Layla wa-layla → Burton, Richard; Weil, Gustav

Al Jallad, Nader. 2011. “A Linguo-Cultural Study of ‘Generosity’ in Arabic.” In: Pamies / Dobrovol’skij (eds.) 2011: 233–239. Arendonk, C. van → van Arendonk, C. al-ʔAʕšā / Geyer, Rudolf. 1905. Zwei Gedichte von al-’A‘šâ / ed., trans., and comm. by R. Geyer. In: Mâ bukâ’u. Vienna, In Kommission bei Alfred Hölder. a passage on maḥāsin al-saḫāʔ (the ‘pros’ of liberality), where a Bedouin / Arab (ʔaʕrābī) and a non-Arab / Persian (ʔaʕǧamī) compete with each other: “[ʔaʕrābī :] ‘We are the most hospitable to our guests.’ – [ʔaʕǧamī :] ‘How that?’ – [ʔaʕrābī :] ‘Because somebody may own nothing but a camel (baʕīr), but as soon as a guest comes he will slaughter it nevertheless.’ Answered the ʔaʕǧamī : ‘We have a better tradition of hospitality than yours.’ – [ʔaʕrābī :] ‘And what would that be?’ – [ʔaʕǧamī :] ‘We call a guest a mehmān, which means that he is the highest [in rank] in the house and the one among us who owns most.’” – ‹Pseudo›-Ǧāḥiẓ [K. alMaḥāsin wa’l-ʔaḍdād, n.d.]: 37. 75 Granara 2010.

Aesthetics of generosity – generous aesthetics

323

al-ʕAskarī, ʔAbū Hilāl al-Ḥasan Ibn ʕAbdallāh. (1964). Ǧamharat al-ʔamṯāl. 2 vols. / ed. ʕAbd al-Maǧīd Qaṭāmiš ... [et al.]. al-Muʔassasa al-ʕarabiyya al-ḥadīṯa liṭ-ṭabʕ wa’n-našr wa’l-tawzīʕ, Cairo. #517: “ʔaǧwadu min Ḥātim”. al-Baġdādī, ʕAbd al-Qādir Ibn ʕUmar. (1967–1969). Ḫizānat al-ʔadab wa-lubb lubāb lisān al-ʕarab / ed. ʕAbd al-Salām Muḥammad Hārūn. Dār al-kātib al-ʕarabī / (later) alHayʔa al-miṣriyya al-ʕāmma li’l-kitāb, Cairo. el-Barghûthî, ʕOmar Ṣáliḥ. 1924. “Rules of Hospitality (Qānûn yḍ-ḍiyâfeh) [sic!].” The Journal of the Palestine Oriental Society Jerusalem, 4: 175–203. Barth, J. 1898. “Zur Kritik und Erklärung des Diwans Ḥâtim Tejjs [sic!].” Zeitschrift der Deutschen Morgenländischen Gesellschaft, 52: 34–74. Bonner, Michael. 2003. “Poverty and Charity in the Rise of Islam.” In: Bonner / Ener / Singer (eds.) 2003: 13–30. Bonner, Michael / Ener, Mine / Singer, Amy (eds.). 2003. Poverty and Charity in Middle Eastern Contexts. State University of New York Press, Albany. Bowen, Donna Lee / Early, Evelyn A. 1993. Everyday Life in the Muslim Middle East. Indiana University Press, Bloomington and Indianapolis. Burton, Richard F. 1885-88. A plain and literal translation of the Arabian Nights’ Entertainments, now entituled The Book of the Thousand Nights and a Night / with an introduction [...] by Richard F. Burton. Burton Club, Benares. Bravman, Meïr M. 1962. “‘The Surplus of Property’: An Early Arab Social Concept.” Der Islam, 38: 28–50. Caussin de Perceval, Armand-Pierre. 1967 [1847–1848]. Essai sur l’histoire des Arabes avant l’islamisme, pendant l’époque de Mahomet, et jusqu’à la réduction de toutes les tribus sous la loi musulmane. 3 vols. Akademische Druck- und Verlagsanstalt, Graz (reprinted from the first edition, Paris: F. Didot, 1847–1848). Chatty, Dawn (ed). 2006. Nomadic Societies in the Middle East and North Africa: Entering the 21st century. Brill, Leiden & Boston. (Handbook of Oriental Studies / Handbuch der Orientalistik, sect. I; vol. 81). Cheikho, Louis → Šayḫū, Luwīs. Chelhod, Joseph. 1990. “Les honneurs du désert.” Le Courrier de l’UNESCO, 43/2: 11– 15. Christes, Johannes. 2006. “Paideia.” Brill’s New Pauly (online). First appeared 2006, retrieved 11 June 2014 from . Derrida, Jacques. 1997. De l’hospitalité / Anne Dufourmantelle invite Jacques Derrida à répondre. Calman-Lévy, Paris. – Translated into English as Of hospitality: Anne Dufourmantelle invites Jacques Derrida to respond, Stanford, Calif.: Stanford University Press, 2000. EI2 = Encyclopaedia of Islam, second edition. Edited by Peri Bearman [et al.]. Leiden: Brill. Brill Online 2014. First published online 2012. El-Enany, Rasheed. 2006. Arab Representations of the Occident: East-West Encounters in Arabic Fiction. Routledge, Abingdon.

324

Stephan Guth

El-Shamy, Hasan M. 1995. Folk Traditions of the Arab World: A Guide to Motif Classification. 2 vols. Indiana University Press, Bloomington and Indianapolis. Farès, B. 2012. “Murūʔa.” In: EI2. Forbes, Duncan. 1880. The Adventures of Hatim Taï: a romance / translated from the Persian. Oriental Translation Fund, London. Freytag, Georg Wilhelm. 1837. Lexicon arabico-latinum praesertim ex Djeuharii Firuzabadiique et aliorum arabum operibus adhibitis Golii quoque et aliorum libris confectum. Vol. 4. C.A. Schwetschke, Halle/Saale. al-Ǧāḥiẓ, ʕAmr b. Baḥr. [n.d.]. K. al-Buḫalāʔ. Dār Ṣādir, Beirut.

al-Ǧāḥiẓ, ʕAmr b. Baḥr. [n.d.]. K. al-Maḥāsin wa’l-ʔaḍdād. . [Probably falsely attributed to al-Ǧāḥiẓ, cf. Rescher (1984)]. [al-Ǧāḥiẓ, ʕAmr b. Baḥr] / “‹Pseudo-› Ǧāḥiẓ”: K. al-Maḥāsin wa’l-masāwī → Rescher (1984). Gershoni, Israel / Jankowski, James P. 1995/2002. Redefining the Egyptian Nation, 19301945. Cambridge University Press, Cambridge [etc]. First published 1995. First paperback edition 2002. (Cambridge Middle East Studies; vol. 2). Geyer, Rudolf. 1905 → al-ʔAʕšā. Geyer, Rudolf. 1898. “Zu den Gedichten des Ḥâtim al-Ṭāʔî.” Wiener Zeitschrift für die Kunde des Morgenlandes, 12: 308–318. Granara, William. 2010. “Nile Crossings: Hospitality and Revenge in Egyptian Rural Narratives.” Journal of Arabic Literature, 41: 121–35. Guth, Stephan. [2015]. “Thus Ruled the Court (Ḥukm al-ʕadāla): A collection of true criminal cases from Syria turned into narratives.” In: Sagaster / Strohmeier (eds.) [2015]: (pagination still not final). Hamori, Andras. 1974. On the Art of Medieval Arabic Literature. Princeton University Press, Princeton, NJ. Hamori, Andras. 2006. “Prudence, Virtue, and Self-respect in Ibn al-Muqaffaʕ.” In: Neuwirth / Islebe (eds.) 2006: 161–180. Ḥātim al-Ṭāʔī. (1963). Dīwān / ed. Karam al-Bustānī. Dār Ṣādir/Dār Bayrūt, Beirut. [Ḥātim al-Ṭāʔī]: Der Dîwân des arabischen Dichters Ḥātim Ṭej / nebst Fragmenten herausgegeben, übersetzt und erläutert von Friedrich Schulthess. Leipzig: J. C. Hinrichs’ sche Buchhandlung, 1897. [Ḥātim al-Ṭāʔī]: The Diwan of Hatim al Tai / translated and commented on by Arthur Wormhoudt. William Penn College, 1984. Hiltbrunner, O. / Gorce, D. / Wehr, H. 1972. “Gastfreundschaft” [Hospitality]. In: RAC VIII: 1061–1123. Ibn Manẓūr, Muḥammad b. Mukarram. (1377 AH [1968 CE]). Lisān al-ʕarab. Dār Ṣādir, Beirut. 15 vols. Ibn Qutayba, ʕAbdallāh Ibn Muslim. (1969). al-Šiʕr wa’l-šuʕarāʔ. 2 parts in 1 vol. – 2 nd ed., Dār al-Ṯaqāfa, Beirut. (First ed. Beirut 1902).

Aesthetics of generosity – generous aesthetics

325

al-ʔIṣbahānī, ʔAbū ’l-Faraǧ. (1323 AH [1905 CE]). Kitāb al-ʔAġānī / ed. ʔA. al-Šinqīṭī. Maṭbaʕat al-Taqaddum, Cairo. – Vol. 16: 93–106 (ʔAḫbār Ḥātim wa-nasabih). Izutsu, Toshihiko. 1966. Ethico-Religious Concepts in the Qurʔān. McGill University Press, Montréal. Jaeger, Werner. 1973. Paideia: die Formung des griechischen Menschen / reprint in 1 vol. of the original edition (in 3 vols.) of 1933–1947. De Gruyter, Berlin [etc.]; Engl. trans. by Gilbert Highet as Paideia: The Ideals of Greek Culture, Oxford: Blackwell, 1939ff. Jallad, Nader Al → Al Jallad, Nader. Kanafani, Aida. [1983] 1993. “Rites of Hospitality and Aesthetics.” In: Bowen / Early 1993: 128–135. – An earlier version appeared in ead., Aesthetics and Ritual in the United Arab Emirates, Beirut: AUB, 1983. Kédros, André. 1990. “L’étranger, l’ami.” Le Courrier de l’UNESCO 43/2: 25–28. Khaleghi-Motlagh, Dj. 1983. “Adab i. Adab in Iran.” Encyclopedia Iranica (ed. Ehsan Yarshater), vol. i, Routledge & Kegan Paul, London [etc.], 431–439. Retrieved from the online edition, (11 June 2014). Leder, Stefan. 2002. “Nomaden und nomadische Lebensformen in arabischer Begrifflichkeit: Eine Annäherung.” In: Leder / Streck (eds.) 2002: 11–40. Leder, Stefan / Kilpatrick, Hilary. 1992. “Classical Arabic Prose Literature: A Researchers’ Sketch Map.” Journal of Arabic Literature, 23/1: 2–26. Leder, Stefan / Streck, Bernhard (eds). 2002. Nomadismus aus der Perspektive der Begrifflichkeit. Proceedings of the 1st meeting in 11 July, 2001. (Orientwissenschaftliche Hefte, vol. 3; Mitteilungen des SFB “Differenz und Integration“, Halle, vol. 1). Lewis, Bernard → Rosenthal, Franz. Lisowski, Georges. 1990. “Une attente d’hôtes.” Le Courrier de l’UNESCO, 43/2: 17–20. Malti-Douglas, Fedwa. 1985. Structures of Avarice: The Bukhalāʔ in Medieval Arabic Literature. Brill, Leiden. (Studies in Arabic Literature, Suppl. to the Journal of Arabic Literature; vol. 11). al-Masʕūdī, ʕAlī b. al-Ḥusayn. (1965-79). Murūǧ al-ḏahab wa-maʕādin al-ǧawhar / ed. Charles Pellat. 7 vols. al-Ǧāmiʕa al-Lubnāniyya, Beirut. (Manšūrāt al-Ǧāmiʕa alLubnāniyya. Qism al-dirāsāt al-tārīḫiyya; vol. 11). al-Maydānī, ʔAḥmad b. Muḥammad. (1284 AH [1867 AH]). K. Maǧmaʕ al-ʔamṯāl. 2 vols. Būlāq. Milinski, Manfred. 2001. “Freigebigkeit lohnt sich.” Spektrum der Wissenschaft, Feb. 2001: 24–26. Monteil, Vincent. 1960. L’arabe moderne. C. Klincksieck, Paris. (Études arabes et islamiques, série 3: Études et documents; vol. 3). Murtaḍā al-Zabīdī, Muḥammad Ibn Muḥammad. (1965 ff.). Tāǧ al-ʕarūs min ǧawāhir alqāmūs. Maṭbaʕat Ḥukūmat al-Kuwayt, al-Kuwayt. (al-Turāṯ al-ʕarabī; vol. 16). Nagel, Tilman. 1981. Staat und Glaubensgemeinschaft im Islam: Geschichte der politischen Ordnungsvorstellungen der Muslime. Vol. II: Vom Spätmittelalter bis zur Neuzeit. Artemis, Zürich & München.

326

Stephan Guth

Nanah, Mouhamed Fouaed. 1987. Freigebigkeit und Geiz in der Vorstellungswelt der vorislamischen arabischen Dichter. Diss. Erlangen-Nürnberg. Neuwirth, Angelika. 2010. Der Koran als Text der Spätantike: ein europäischer Zugang. Verlag der Weltreligionen, Berlin. Neuwirth, Angelika / Islebe, Andreas Christian (eds.). 2006. Reflections on Reflections: Near Eastern Writers Reading Literature. Reichert, Wiesbaden. (Literaturen im Kontext; vol. 23). Nippa, Annegret. 2006. “Art and Generosity: Thoughts on the Aesthetic Perceptions of the ʕArab.” In: Chatty (ed). 2006: 539–567. Orel, Vladimir E. / Stolbova, Olga V. 1994. Hamito-Semitic Etymological Dictionary: materials for a reconstruction. Brill, Leiden. (Handbuch der Orientalistik, 1. Abteilung: Der Nahe und Mittlere Osten; vol. 18). Pamies, Antonio / Dobrovol’skij, Dmitrij (eds.). 2011. Linguo-Cultural Competence and Phraseological Motivation. Schneider-Verlag Hohengehren, Baltmannsweiler. Paret, Rudi. 1980. Der Koran. Vol. II: Kommentar und Konkordanz. Kohlhammer, Stuttgart [etc.]. Propp, Vladimir. [1928] 1968. Morphology of the Folktale / transl. by Laurence Scott [and] with an introd. by Svatava Pirkova-Jakobson. University of Texas Press, Austin. – Russian original: Morfologija skazki, Leningrad: Academia, 1928. RAC = Reallexikon für Antike und Christentum: Sachwörterbuch zur Auseinandersetzung des Christentums mit der Antiken Welt / in Verbindung mit Carsten Colpe [et al.] hg. von Theodor Klauser. Vol. VIII: Fluchtafel (Defixion) – Gebet I. Stuttgart: Anton Hiersemann, 1972. Rebhan, Helga. 1986. Geschichte und Funktion einiger politischer Termini im Arabischen des 19. Jahrhunderts (1798-1882). Harrassowitz, Wiesbaden. Rescher, Oskar. 1925. Abriss der arabischen Litteraturgeschichte. I. Band. [s.n.], Stuttgart. Rescher, Oskar / ‹Pseudo-› Ǧāḥiẓ. (1984). Das kitâb al-mahâsin [sic!] wa’l-masâwî (Ueber die guten und die schlechten Seiten der Dinge / nach G. van Vlotens Ausgabe (Leyden 1898) aus dem Arab. übers. von Oskar Rescher. 2 parts in 1. Biblio-Verlag, Osnabrück. (part 1: reprint of the first edition, Constantinople 1926; part 2: reprint of the first edition, Stuttgart 1922). (Oskar Rescher, Gesammelte Werke; Abt. 2: Schriften zur Adab-Literatur; vol. 3). Richard, Yann. 1990. “Au pays des mille et une politesses.” Le Courrier de l’UNESCO, 43/2: 30–33. Rolland, Jean-Claude. 2014. Etymologie arabe: Dictionnaire des mots de l’arabe moderne d’origine non sémitique. Éditions Lulu.com. Rosenthal, Franz / Lewis, Bernard. [2012]. “Ḥurriyya.” In: EI2. Brill online (retrieved 12 Dec., 2011). Rückert, Friedrich. 2004 [1846]. Hamâsa, oder Die ältesten arabischen Volkslieder / gesammelt von Abu Temmâm; übersetzt und erläutert von Friedrich Rückert. Olms, Hildesheim. (Reprint in 1 vol. of the original edition, Liesching, Stuttgart, in 2 vols.).

Aesthetics of generosity – generous aesthetics

327

Sagaster, Börte / Strohmeier, Martin (eds.). [in preparation for publication in 2014]. Crime Fiction in and around the Eastern Mediterranean. Proceedings of the workshop held in Nicosia/Cyprus, 11-12 November, 2011. Harrassowitz, Wiesbaden. (mîzân; vol. 23). Šayḫū, Luwīs [Cheikho, Louis]. 1890. K. Šuʕarāʔ al-Naṣrāniyya. Ǧuzʔ 1: šuʕarāʔ alǧāhiliyya. Maṭbaʕat al-ʔābāʔ al-mursalīn al-yasūʕiyyīn, Beirut. Schulthess, Friedrich. 1897. → [Ḥātim al-Ṭāʔī]: Der Dîwân des arabischen Dichters Ḥātim Ṭej. Shamy, Hasan M. El → El-Shamy, Hasan M. Sharlet, Jocelyn. 2011. “Tokens of Resentment: Medieval Arabic Narratives About Gift Exchange and Social Conflict.” Journal of Arabic and Islamic Studies, 11: 62–100. Shryock, Andrew. 2004. “The New Jordanian Hospitality: house, host, and guest in the culture of public display.” Comparative Studies in Society and History, 46/1: 35–62. al-Tanūḫī, al-Qāḍī. al-Mustaǧād min fiʕalāt al-ʔaǧwād. . al-Ṭahṭāwī, Rifāʕa Rāfiʕ. [1993]. Taḫlīṣ al-ʔibrīz fī talḫīṣ Bārīz. 3 vols. al-Hayʔa almiṣriyya al-ʕāmma lil-kitāb, Cairo. al-Ṭahṭāwī, Rifāʕa Rāfiʕ. 2011. An Imam in Paris: Account of a Stay in France by an Egyptian Cleric (1826-1831): (Takhlīṣ al-Ibrīz fī Talkhīṣ Bārīz aw al-Dīwān al-Nafīs biĪwān Bārīs) / introduced and translated by Daniel L. Newman. Saqi, London. (First published in hardback in 2004 by Saqi Books.) el-Tayib, Abdullah. [ca. 1962]. “Themes of Hospitality in Arabic Poetry.” Kano Studies, [no. ?]: 16–32. The Thousand and One Nights → Burton, Richard; Weil, Gustav. Thouvenin, Georges. 1933. “La légende arabe d’Hatim Ta’ï dans le Décaméron.” Romania, 59: 247–269. van Arendonk, C. 2011. “Ḥātim al-Ṭāʔī”. EI2. Brill online (retrieved 12 Dec., 2011) Wagner, Ewald. 1987. Grundzüge der klassischen arabischen Dichtung. Vol. I: Die altarabische Dichtung. Vol. 2: Die arabische Dichtung in islamischer Zeit. Wissenschaftliche Buchgesellschaft, Darmstadt. (Grundzüge; vols. 68 and 70). Weil, Gustav. 1838-41. Tausend und eine Nacht: Arabische Erzählungen / zum ersten Male aus dem arabischen Urtext treu übersetzt von Dr. Gustav Weil. 4 vols. Verlag der Classiker, Stuttgart (vol. I); Dennig, Finck & Co., Pforzheim (vols. II÷IV). Weir, T.H. 1994. “Ṣadaḳa”. EI2, vol. viii: 708–716. Wielandt, Rotraud. 1980. Das Bild der Europäer in der modernen arabischen Erzähl- und Theaterliteratur. Orient-Institut, Beirut / Steiner, Wiesbaden. (Beiruter Texte und Studien; vol. 23). Wormhoudt, Arthur. 1984. → [Ḥātim aṭ-Ṭāʔī]: The Diwan of Hatim al Tai. al-Zabīdī → Murtaḍā al-Zabīdī. al-Zamaḫšarī, Maḥmūd b. ʕUmar. [n.d.]. Rabīʕ al-ʔabrār. (retrieved 01 July, 2014). Zysow, A. 2001. “Zakāt”. EI2, vol. xi: 406–422.

Who was Ḥubbā al-Madīniyya? Pernilla Myrne, University of Gothenburg

Classical Arabic literature features many stories about a woman named Ḥubbā al-Madīniyya. These stories are mostly humorous and often lewd. Some readers describe them as overly obscene confirmation of early Arabic high literature’s tendency to degrade women. One modern scholar has claimed that Ḥubbā “embodies in her essence the female adab [classical Arabic belles-lettres] character”, in that she is, first and foremost, a body with “uncontrollable sexuality”.1 Other scholars counter this interpretation, arguing that it is predicated upon a misreading of classical Arabic literature and a lack of understanding of its genres.2 The focus here is the literary creation of a quasi-historical character, and its variation across genres. Behind the layers of myth, however, there may be a true lived experience. That conflicting accounts exist does not rule out seeking an answer to the question, who was Ḥubbā al-Madīniyya? Ḥubbā in ʾadab Jāḥiẓ (d. 868) made Ḥubbā famous – or, rather, infamous – by mentioning her in his book on animals, Ḥayawān, and in a humorous epistle, Mufāḫarat al-jawārī wa-l-ġilmān (The boasting match between girls and boys). According to Jāḥiẓ, Ḥubbā was married to a man named ʾAbū Kilāb ibn Lisān al-Ḥummara, also called Ibn ʾUmm Kilāb.3 Jāḥiẓ quotes a poem attributed to Hudba ibn Ḫašram al-ʿUḏrī (see below), which suggests that Ḥubbā was attracted to the youth and appearance of Ibn ʾUmm Kilāb, but did not love him truly: No mother has loved the way I love her [My love is] not like Ḥubbā’s love for Ibn ʾUmm Kilāb

1

Malti-Douglas, Women’s Body, 45–46.

2

Scott Meisami, “Reading medieval women”.

3

Jāḥiẓ, Ḥayawān, vol. 2, 200.

Who was Ḥubbā al-Madīniyya?

329

She saw him with long, beautiful forearms Perhaps she was revived [inbaʿaṯat] by strength and youth.4 Jāḥiẓ also quotes a poem that he attributes to Ḥubbā: I wish that he was a lizard and I would be like a lizardess in wasteland [kudyatin5] that had found a vacant space. Jāḥiẓ claims that she recited this poem for her son, who reproached her for having married Ibn ʾUmm Kilāb – a young man – when she was almost fifty.6 The poem appears in the section on lizards in Ḥayawān and, surprisingly, is there to illustrate the poet’s point that lizards have two sets of genitals.7 Ḥubbā’s poem shows, according to Jāḥiẓ, that “she wished that she had two vaginas and he had two penises.”8 This interpretation is not obvious; but for Jāḥiẓ, Ḥubbā is a comical character. Al-Jāḥiẓ uses Ḥubbā to colour his arguments and to embody lively elements of entertainment. In Mufāḫarat al-jawārī wa-l-ġilmān, Jāḥiẓ relates an anecdote about Ḥubbā scaring away the camels of the caliph ʿUṯmān ibn ʿAffān.9 She is one of the lustful women (al-muġtalimāt), al-Jāḥiẓ writes, whom women from Medina consult about women’s groaning (qabʿ) while having sex. They ask if it is an old habit, or something the women – presumably the women of Medina – invented. Ḥubbā responds: My daughters, I went on the minor pilgrimage with the Commander of the Faithful, ʿUthmān, May God be pleased with him. As we were returning and came to al-ʿArj10, my husband looked at me and I at him. He was attracted by what he saw and so was I and so he fell upon me. ʿUṯmān’s camels were passing by as I was groaning loudly, having being overtaken by that which happens to the daughters of Adam, whereupon the camels ran away. There were five hundred camels, and none of them have yet been found.11

4 Ibid., 200–201. 5 A piece of land with hard soil, impossible to cultivate; Ibn Manẓūr, Lisān, s.v. k-d-y. 6 She was above naṣaf, which means, according to the editor, 45 or 50 years; Ḥayawān, vol. 6, 75, n. 2. 7 Ibid., 72–73. 8 Ibid., 75. Quoted by Yāqūt, Muʿjam, 1003. 9 Jāḥiẓ, Rasāʾil, 2, 129. 10 A place between Mecca and Medina. 11 Jāḥiẓ, Rasāʾil, 2, 129-30.

330

Pernilla Myrne

Ḥubbā educates the women of Medina not only about groaning, but also about moving (ġarbala) during sexual intercourse.12 The curious words explained by Jāḥiẓ – qabʿ and ġarbala – lend an air of exoticism to the story. What is more, al-Jāḥiẓ continues, when Ḥubbā’s own son asks her how to please a woman, she offers him advice as well.13 These incidents are retold correspondingly by Ibn Abī Ṭāhir Tayfūr (d. 893), al-Jāḥiz’s somewhat younger contemporary, in Balaġāt al-nisāʾ (Instances of women’s eloquency), one of the surviving volumes of his Kitāb al-manṯūr wa-l-manẓūm. There are some differences, however. In Balaġāt al-nisāʾ, Ḥubbā recounts the story of the caliph’s camels to her own daughter, who complains about her husband asking her to groan when they have sex. Furthermore, the story about Ḥubbā’s son getting sexual advice from his mother is elaborated upon. Here, “young men from Quraysh” were discussing how to please a woman, and urged Ḥubbā’s son to ask his mother about it.14 Ibn Abī Ṭāhir tells some additional stories highlighting Ḥubbā’s lustfulness and licentiousness that are discussed below. Jāḥiẓ’s epistle is an entirely joking debate between an owner of slave boys (ġilmān) and an owner of slave girls (jawārī). The subject is whether slave girls or slave boys are preferred as sexual partners.15 The second part of the epistle, where we find the anecdotes about Ḥubbā, contains numerous amusing anecdotes with the explicit aim of entertaining the reader – including stories about sodomites, effeminates, and various anonymous and named men and women.16 Ibn Abī Ṭāhir’s volume on eloquent women is generally more serious. It starts by quoting speeches from ʿĀʾisha bint ʾAbī Bakr and other famous early Muslim women. The anecdotes about Ḥubbā, however, appear in a chapter devoted to mujūn; that is, vulgar, indecent talk intended to be humorous. Contrary to Jāḥiẓ, he introduces his anecdotes with a chain of transmitters (isnād), with claims of authenticity – although it is unlikely that they were received as anything more than entertainment.

12 Ibid., 130. 13 Ibid., 131. 14 Ibn ʾAbī Ṭāhir, Balaġāt, 194. There are other differences. For example, in Ibn Abī Ṭāhir’s version, Ḥubbā’s husband is not mentioned and the camels were eventually found. See the translation of the anecdotes about Ḥubbā in Borg, “Obscenities,” 152–153. 15 Cf. Rosenthal, “Male and female”. 16 Jāḥiẓ, Rasāʾil, 125.

Who was Ḥubbā al-Madīniyya?

331

The chain of transmitters is not reliable. Ibn ʾAbī Ṭāhir attributes the anecdotes to al-Hayṯam ibn ʿAdī, who, in turn, received some of them from Ṣāliḥ ibn Ḥassān. Ṣāliḥ came from Medina, held a poetry majlis in Kufa, and died at the time of the caliph al-Mahdī (r. 775–785), a century or so after Ḥubbā, who, as we have seen, presumably lived at the time of Caliph ʿUṯmān ibn ʿAffān (r. 644–656).17 Ṣāliḥ was also a ḥadīṯ transmitter, but was dismissed as such by al-Nasāʾī, al-Buḫārī, and others. In the words of Ibn Ḥibbān, “He used to have female singers and listen to music, and he used to narrate fabricated reports on the authority of trustworthy narrators.”18 Ṣāliḥ’s pupil, al-Hayṯam ibn ʿAdī alṬāʾī (d. 821), was also considered unreliable among traditionalists.19 He was born in Kufa and met with some success at the ʿAbbāsid court in Baghdad. Charles Pellat suggests that although Jāḥiẓ quotes al-Hayṯam ibn ʿAdī quite often, he “seems to consider him proverbial for the lack of authenticity of his sources.”20 Yet, based on precisely the unreliable anecdotes related by Jāḥiẓ, Pellat does not hesitate to call Ḥubbā one of the “true professionals of love”, which is likely a euphemism for a prostitute.21 Perhaps Pellat intended the epithet “true professionals of love” to portray Ḥubbā as a devoted sexual adviser, as both Jāḥiẓ and Ibn ʾAbī Ṭāhir claim that she gave sexual advice to both women and men. The only man she had sex with – according to these and other litterateurs – was her husband, so it is of course farfetched to deduce that she was a prostitute. In her book about misogynistic attitudes in Classical Arabic literature, Woman’s body, woman’s word, Fedwa Malti-Douglas adopts yet another approach to Ḥubbā. Malti-Douglas thesis is that these texts defined women foremost as bodily creatures, whose ability to speak was intrinsically linked to their bodily – especially sexual – functions, whereas men were considered in terms of their mental capacities and skills. Consequently, women in classical Arabic belleslettres (ʾadab) talked mostly about sex and used their bodies to gain opportunities to speak. According to Malti-Douglas, Ḥubba is the prototype for “the fe-

17 For Ṣāliḥ ibn Ḥassān as one of al-Hayṯam ibn ʿAdī’s authorities, see Leder, Korpus, 50–51. 18 Quoted in al-ʾAlbānī, Salah, 146. 19 Pellat, ”Al-Hayṯam b. ʿAdī”. 20 Pellat, ”Al-Hayṯam b. ʿAdī”. 21 Pellat, ”Djins”.

332

Pernilla Myrne

male adab character.”22 Her “biological function is tied to her discursive function” and, accordingly, women represented by Ḥubbā embody “uncontrollable sexuality that expresses itself not only through actions but also through words.”23 Malti-Douglas even argues that Ḥubba is the archetypical woman in Classical Arabic literature, a characterization based on the fact that al-Maydānī (a twelfth-century writer) claims that the women in Medina called her Eve, the mother of mankind (see below). From this depiction, Malti-Douglas deduces that Ḥubbā has fused with Eve, or, as she puts it, she “becomes Eve, the archetypical female.”24 The erotic Ḥubbā Ḥubbā’s “uncontrollable sexuality” is best exemplified by two anecdotes in Balaġāt. Malti-Douglas mentions one of them, but probably also had the other in mind when she refers to Ḥubbā’s “overtly (and overly) sexual nature.”25 The two anecdotes are told on the authority of al-Hayṯam ibn ʿAdī and Ṣāliḥ ibn Ḥassān and portray Ḥubbā as outrageously libidinous.26 In the first, she tells the “girls of Quraysh” that she satisfied herself with her grandson’s dog – to which the girls respond that she should not be ashamed, assuring her, “it is a noble deed.”27 In the second, she asks her three daughters to tell her their most exciting sexual experience. She dismisses the accounts of the two oldest for being too dull, but when the youngest daughter starts to tell her a sexually explicit account, she screams, “Shut up, daughter, shut up immediately, your mother is wetting herself from passion.”28 These stories are, indeed, “scabrous, obscene and to a certain extent tasteless”, as Gert Borg has described them.29 In fact, this characterization is quite in order, as the stories belong to the genre of mujūn.30

22 Malti-Douglas, Woman’s body, 45. 23 Ibid., 46-47. 24 Malti-Douglas, Women’s body, 46. 25 Ibid., 47. 26 Ibn Abī Ṭāhir, Balaġāt, 195. Translated by Borg, ”Obscenities”, 153. 27 See Borg, “Obscenities,” 153. 28 The translation in ibid. 29 Ibid., 149. 30 See Scott Meisami’s criticism of Malti-Douglas’ treatment of the comical anecdotes about Ḥubbā; “Writing Medieval Women,” 60–61. One of her points is that mujūn is meant to be “humorous rather than pejorative,” ibid. 61.

Who was Ḥubbā al-Madīniyya?

333

Probably few who read or heard the story of Ḥubbā having sex with a dog really believed that it had happened – or, yet more incredible, that the Qurayshite girls in her company considered it “a noble deed”. Nevertheless, many readers possibly found it hilarious to think that they did. They may well have found Ḥubbā’s “uncontrollable sexuality” exceedingly amusing as well, as sexual comedy is a cherished mujūn topic. But it is inaccurate to claim that this particular kind of literature degrades women. In fact, anecdotes of aberrant sexuality are even more common with male protagonists in such stories. Sex with animals seems to have been a fairly common licentious theme, typically involving men and dogs or mules.31 Scores of other medieval stories would likely shock a twenty-first century reader, and most of them are about men. Irreverent stories about Ḥubbā, were, as we have seen, first told by Ṣāliḥ ibn al-Ḥassān and Hayṯam ibn ʿAdī in early Abbasid Kufa and Baghdad. But these versions were elaborated upon; in the earliest extant erotic compilation, Jawāmiʿ al-laḏḏa, Ḥubbā is a full-fledged sexual authority and adviser. This book was probably written in late ninth-century Baghdad. The author, ʿAlī ibn Naṣr, often quotes al-Jāḥiẓ and Ibn ʾAbī Ṭāhir; for example, recounting the anecdote about Ḥubbā and the caliph’s camels.32 His main material on Ḥubbā, however, must have derived from other sources. She acts as an expert, offering opinions about groaning and men’s and women’s preferences – such as women’s preference for beardless men.33 Notably, Ḥubbā is an ardent heterosexual. Indeed, in a chapter devoted to lesbianism, she warns women against thinking that they can be sexually satisfied without men.34 One long story about her stands out.35 It is faithful to the theme in the older anecdotes in which she gives advice to her son and talks frankly about sex with her daughters. In Jawāmiʿ al-laḏḏa, she offers detailed advice to her daughter on her wedding day about how to arouse her husband and satisfy him, how to behave during their sexual union, and what to say. Thereafter, she proceeds to give her son-in-law advice about how to please and satisfy his wife. Ibn Naṣr presents her as one of the ancients (al-qudamāʾ) and Bedouins, comparing her frank conversation to that of modern and settled people, which is much less explicit. 31 Cf. Munajjid, al-Ḥayāt al- jinsiyya, 93–95. 32 Ibn Naṣr, ms. Aya Sofya 3836 fols. 138a and 87a. 33 Ibid., 137a–138a, and e.g., 79a. 34 Ibid., 87a. 35 Ibid., ms. Aya Sofya 3837, fols. 41a–44a.

334

Pernilla Myrne

Ḥubbā in history writing Ḥubbā may be a mujūn phantasy, but she also claims a place in history books, where we find a very different Ḥubba – although reports of her are scarce and do not give much information. She is mentioned by al-Balāḏurī (d. 892) and alṬabarī (d. 923). Al-Balāḏurī mentions her twice. On the first occasion, he relies on al-ʿUmarī, who quotes al-Hayṯam ibn ʿAdī, the man behind more licentious stories about Ḥubbā.36 In this instance, he claims that Ḥubbā helped Muṣʿab ibn al-Zubayr (d. 691) find a wife. The candidate she presents to him was the aristocratic ʿĀʾisha bint Ṭalḥa, whom he marries. Any woman chosen for the task of matchmaking was likely considered reliable and possessing good taste. Moreover, she would have to have had access to aristocratic women and be allowed to meet with men other than her relatives. According to Kitāb al-ʾaġānī, this particular matchmaking was executed by the singer ʿAzza al-Maylā.37 Another report by al-Balāḏurī is more intriguing. He relates:38 ʿAbd al-Malik went on the pilgrimage and passed by the house of Ḥubbā al-Madīniyya in Medina. The young men (fityān) of Quraysh sat and talked with her. She looked down at him and he looked at her. Then she called out to him and he got up. He said, “Ḥubbā, I am ʿAbd al-Malik.” She answered, “I know that, may you be ransomed with my father and mother. Thanks to God who let me see your face before I die. How are you, my lord?” He said, “I am fine, Ḥubbā, how is your chilled water and those of the young men of Quraysh who are visiting you?” She answered, “They are fine, Commander of the faithful. You killed your brother, ʿAmr ibn Saʿīd.” He said, “Yes, by God. It pains me, but he wanted my death.” She said, “May he not rise again!” He gave order to give her five hundred dinar. She in turn gave him gifts, and he kissed her. The caliph ʿAbd al-Malik killed his relative ʿAmr ibn Saʿīd in 689, when the latter tried to take over the government. This event transpired during a turbulent time, especially for the people in Hijāz. The appointment of Yazīd I as caliph in 680 was unpopular among the people of Medina.39 The same year, Ḥusayn ibn ʿAlī left Medina with his family, and was killed in Karbalāʾ some months later; the women were taken prisoners. Shortly thereafter, ʿAbdullāh 36 Al-Balāḏurī, ʾAnsāb, vol. 5, 203. 37 ʾAbū l-Faraj, ʾAġānī, vol. 11, 122–124. 38 Told on the authority of ʿUmar ibn Bakīr, Hišām ibn al-Kalbī and Ibn Miskīn alMadīnī; Balāḏurī, ʾAnsāb, vol. 11, 186. 39 Cf. Buḫāri, vol. 6, book 60, ḥadīṯ 352.

Who was Ḥubbā al-Madīniyya?

335

ibn al-Zubayr started a revolt against the Umayyad rule, supported by the people of Medina, and was declared caliph by his supporters in 683, after the death of Yazīd I. ʿAbdullāh, seated in Mecca, was defeated by the caliph ʿAbd al-Malik in 692. When ʿAbd al-Malik conducted this pilgrimage, ʿAbdullāh was still in charge of Mecca. It must have been crucial for ʿAbd al-Malik to know what the young Qurayshite men in Medina though about his leadership, so he went to the place where he knew he could get information: Ḥubbā’s quite warm welcome must have calmed him somewhat. Whether this report is true or not, it suggests that Ḥubbā had a distinct standing in Medina, which was well-known outside the city. Next time we meet Ḥubbā in the history books is in 691, after the killing of Muṣʿab ibn al-Zubayr, ʿAbdallāh’s brother. ʿAbdallāh had appointed him governor of Basra, and he was popular among the people of Medina. Ḥubbā had also been personally involved with him – at least according to the report by alBalāḏurī. Al-Ṭabarī recounts that when ʿAbd al-Malik made the pilgrimage in 691, “Ḥubbā approached him and said, ‘Did you kill your fellow tribesman Muṣʿab?’ He replied: Whoever tastes war, finds its taste / bitter, and it leaves him in a rough country.”40 Al-Ṭabarī says nothing more about Ḥubbā; but even if this incident does not reveal who she was, read together with the incident recounted by al-Balāḏurī, it suggests that she held a position of influence. Fictional character or no, she incorporates the feelings of Medina, or at least those of the fityān, in these historical texts. ʿAbd al-Malik is eager to know her sympathies after he killed ʿAmr. After he had killed Muṣʿab, however, he has strayed too far from Ḥubbā’s point of view, and she confronts him directly. In addition to the reports by al-Balāḏurī and al-Ṭabarī, Ḥubbā is mentioned in connection to the poet Hudba ibn Ḫašram al-ʿUḏrī and the poem quoted in the beginning of this essay. Hudba was a famous poet in Hijaz in the early Umayyad era. He killed his relative in a dispute, and was executed in 674, after some years in prison.41 Al-Mubarrad (d. 898) and ʾAbū l-Faraj al-ʾIṣfahānī (d. 967) claim that he recited the poem as he was walking toward the site of his

40 Translated by Michael Fishbein in Ṭabarī, History, vol. 21, 187. 41 Sezgin, GAS, vol. 2, 265. ʾAbū l-Faraj al-ʾIsfahānī devotes one article on him in ʾAġānī, vol. 21, 179-92.

336

Pernilla Myrne

execution, passing Ḥubbā along the way.42 Ḥubbā purportedly accused him of neglecting his wife, who was very beautiful – according to our chroniclers – and shed many tears for her ill-fated husband. When Hudba passes Ḥubbā on the way to the execution site, his wife is walking behind him, sobbing. The story in ʾAġānī goes: Hubda walked past Ḥubbā and she said to him, “I counted you as one of the fityān, but I renounce you today, because I do not deny that men should meet death with serenity (yaṣbiru ʿalā l-mawt), but how can you refrain from (yaṣbiru ʿan) this woman?” He answered, “By God, Ḥubbā, she has a strong [case/verse?]. If you want, I will describe it for you.” The people stopped, and he recited: “I have loved her ardently in a way that no mother could love/ Not like Ḥubbā’s love for Ibn ʾUmm Kilāb She saw him with long, beautiful forearms/ Just as the strength and youth that she desired” After his poem, she turned around and fled in to her house, locking the door behind her. This story suggest that Hudba used to be one of the fityān who visited Ḥubbā, and that a real member of this group would never chose death to save his honour over life with a beautiful woman. If it is correct, this interpretation contradicts reports about his execution that maintain he tried to be pardoned and thus escape the death penalty. He even asked the prophet’s wife, ʿĀʾisha bint ʾAbī Bakr, to help him, but the relatives of his victim refused to pardon him. In Mubarrad’s version of the story, Ḥubbā makes a somewhat stronger case. When Hudba is escorted to the execution site, he starts to recite poetry. Ḥubbā then says to him, “I have never seen someone with a harder heart than you. How can you recite poetry when you are walking towards your death, with her behind you, mourning like a thirsty gazelle?”43 So who was Ibn ʾUmm Kilāb, whom Hudba in this story apparently accuses Ḥubbā of not having loved enough? The man suggested by Jāḥiẓ, Ibn Lisān alḤummara, is mentioned in connection to the caliph Muʿāwiya (r. 661–680), and was known for his “wisdom, his eloquence, his gift for lively repartee and his

42 Al-Mubarrad, Fāḍil, 766–767; ʾAbū l-Faraj, ʾAġānī, vol. 21, 188 and 190. The chain of transmitters in ʾAġānī is Ḥammād ibn Isḥāq – his father – Ibn Kunnāsa. 43 Mubarrad, Kāmil, 766–767.

Who was Ḥubbā al-Madīniyya?

337

profound knowledge of men and women”.44 I have not found that he is called by the nasab Ibn ʾUmm Kilāb, however. The history books mention another man with this nasab, namely an Ibn ʾUmm Kilāb, who informed ʿĀʾisha that the people in Kufa had pledged allegiance to ʿAlī after the murder of caliph ʿUṯmān ibn ʿAffān, in 656.45 His name was ʿUbayd ibn ʾAbū Salama al-Layṯī, according to Ibn Saʿd, who mentions him briefly in his compendium of companions and successors.46 Ḥubbā also appears as a character in a book that contains rather long quasihistorical tales, namely al-Maḥāsin wa-l-ʾaḍdād, by an unknown writer.47 Here, she is called Ḥawwāʾ, that is Eve. The book was written no earlier than the middle of the tenth century, so it may be contemporary with Jawāmiʿ al-laḏḏa.48 The author relates the story of a woman in Medina called Ḥawwā’, who is obviously our Ḥubbā, attributing the story to Ṣāliḥ ibn Ḥassān, the transmitter in Ibn ʾAbī Ṭāhir’s Balaġāt:  49 I saw a woman in Medina who was called Ḥawwā’. She was the one who taught the women in Medina al-naqʿ, which is snoring and moving, alġarbala, and al-rahz.50 She had a porch with a roof [saqīfa] where important men from Quraysh used to talk to her. Every family in Medina let their boys suck her breast, or one of her daughters’ breasts. The people in Medina called her Ḥawwā. Every noble man in Medina who had sat in her shelter sent her thirty loads or more each year of food, dates, money, servants and garments. One day, Ṣāliḥ narrates, according to the author of Maḥāsin, Muṣʿab ibn alZubayr, ʿAmr ibn Saʿīd ibn al-ʿĀṣ, and a son of ʿAbd al-Raḥmān ibn ʾAbī Bakr come to her and ask her to have a look at the women they are about to marry. They had proposed to three aristocratic women, but could not gain access to them before the wedding night. Muṣʿab’s fiancée was, of course, ʿĀʾisha bint 44 Pellat, ”Ibn Lisān al-Ḥummara”. 45 Balāḏurī, ʾAnsāb, vol. 2, 217–218, vol. 5, 91. Ṭabarī, vol. 16, 38. 46 Ibn Saʿd, Ṭabaqāt, vol. 5, 65. 47 The book was attributed to Jāḥiẓ early on, but this attribution is now definitely refuted, see Pellat, “Nouvel essai”, 147. 48 Parts of the book rely on al-Maḥāsin wa-l-masāwī by al-Bayhaqī (3rd-4th/9th-10th century); see Gériès, Maḥāsin, and Gériès, Bayhaqī. 49 Ps.-Jāḥiẓ, Maḥāsin, 221. 50 According to Jāḥiẓ, this two words have the same meaning, i.e. movements in sexual intercourse.

338

Pernilla Myrne

Ṭalḥa, whom we met before. Ḥubbā goes to see them and asks them to undress and show all parts of their bodies to her, which she is then able to describe eloquently to the men. This story in Maḥāsin is, in fact, a blend of several earlier anecdotes. Ḥawwāʾ, Eve, may be a misreading of Ḥubbā. She teaches the women the exotic activities al-naqʿ, which must be al-Jāḥiẓ’s al-qab, al-ġarbala, and al-rahz, which are all mentioned by Jāḥiẓ. The incident with Muṣʿab and his two noble companions is, in fact, found in ʾAġānī – but in that version, the singer ʿAzza alMaylāʾ gets to look at the three aristocratic women. The story in ʾAġānī has the same narrator, Ṣāliḥ ibn Ḥassān, who relates that ʿAzza was very popular in Medina and used to sit with noble men, just as Ḥubbā is depicted in Maḥāsin. Al-Balāḏurī’s account of Ḥubbā acting as matchmaker to Muṣʿab is much more sober. The story about Ḥawwāʾ in Maḥāsin is obviously fictional, enhanced by the imaginative details provided by the storyteller, which present her as immensely popular and rich. Proverbial Ḥubba In addition to historical reports and stories, traces of Ḥubbā as a wise woman show up in early Arabic literature. She appears as a sage, with answers to all sorts of (non-sexual) questions. Ibn ʾAbī Ṭāhir quotes Ḥubbā’s more decorous sayings in a section on judicious and eloquent women, which is virtually the opposite of that of mujūn.51 Ibn ʾAbī Ṭāhir writes on the authority of Isḥāq alMawṣilī that, “Ḥubbā al-Madīniyya was asked, ‘What is the wound that cannot be healed?’ She answers, “The noble man’s need for the wretched man, and then he does not repay him.” A similar saying is attributed to her in ʿUyūn alʾaḫbār by Ibn Qutayba.52 She is also asked about the nature of honour and humiliation, which she responds to in the same concise manner. Ḥubbā’s image as a wise woman, however, has not become famous. The most common later image of her, I believe, is the one spread by al-Maydānī (d. 1124) in his collection of proverbs. Malti-Douglas refers to this book when she discusses Ḥubbā, and so does al-Munajjid. Al-Maydānī has his version from earlier collections of proverbs, and possibly from other sources as well. Hamza al-ʾIṣfahānī (d. after 350/961) first mentioned her in his collection of proverbs,

51 Ibn ʾAbī Ṭāhir, Balaġāt, 176. 52 Ibn Qutayba, ʿUyūn, vol. 3, 156.

Who was Ḥubbā al-Madīniyya?

339

ʾAmṯāl ʿalā ʾafʿāl, in the saying “Hornier than Ḥubbā” (ʾašbaqu min Ḥubbā).53 Hamza al-ʾIṣfahānī was born in Isfahan around 893 and lived there all his life, with the exception of a few visits to Baghdad.54 His collection of proverbs is the first to deal with comparative proverbs, and was used by the later collectors ʾAbū Hilāl al-ʿAskarī and the more famous al-Maydānī. The following is Hamza’s version: She was a woman from Medina who was married several times. At old age, she married a young man who was called Ibn ʾUmm Kilāb. Her middleaged son got up and walked to Marwān ibn al-Ḥakam, who was governor of Medina.55 He said, “My shameless mother married a youngster in the prime of youth and of young age, in spite of her and my age. She has made me and herself a subject of gossip.” Marwān asked her and her son to come to him, which she did. She did not, however, pay attention to his talk, instead she turned to her son, and said: “You son of the donkey saddle, have you seen that tall young man? By God, let him throw down your mother right in the doorway, so that he may cure her ardent desire and may her soul go forth below him. I wish that he was a lizard and that I was his little friend56, then we would have found a vacant place. Her words were spread and she became the subject of proverbs. One of those who made an example of her is Hudba ibn Ḫašram. He said, “No mother loved passionately like my love for her / Not Ḥubbā’s love for Ibn ʾUmm Kilāb / She saw him tall and with bulky forearms / perhaps she was revived [inbaʿaṯat] by strength and youth. In the printed version, the section on Ḥubbā is longer, but the editor notes that the extended version is omitted in three of the four manuscripts he used for the edition. In the extension, it is claimed that the women of Medīna called Ḥubbā Eve (Ḥawwāʾ), the mother of mankind, as she taught them different types of coition. This is the name given to Ḥubbā in the quasi-historical Maḥāsin, as we have seen. We can be quite certain that the extension is a later addition. It is a verbatim copy of al-Maydānī’s version, and must have been added by a copyist who compared the two authors. Al-ʿAskarī’s version is short, like Hamza’s version quoted above, but with some variances. For example, the poem is attributed to Ibn Harma instead of Hudba, which alters the interpretation of it, as Ibn Harma was born in 709, when Ḥubbā probably was 53 Al-ʿAskarī, Jamhara, vol. 1, 562–563. 54 Rosenthal, ”Ḥamza al-Iṣfahānī”. 55 Marwān was governor of Medina under Muʿāwiya I (r. 661-680). 56 ḍabb is lizard; here: ḍubayba, which I have chosen to translate as “little friend”.

340

Pernilla Myrne

no longer alive.57 In fact, al-ʿAskarī got much of his material from his relative ʾAbū ʾAḥmad al-ʿAskarī (ca 906-993), who studied in Isfahan and may have met Hamza al-ʾIsfahānī there. If Hamza al-ʾIṣfahānī had really written the longer version, al-ʿAskarī would probably have quoted some of it. The version of al-Maydānī, which would become the most widespread account, seems to be a synthesis of various licentious traditions. For example, he has Ḥubbā’s daughter saying that her husband “filled her house with wealth and her cunt with prick”, which is a saying otherwise attributed to ʿUmar ibn ʿUbaydallāh al-Taymī, when he asked ʿĀʾisha bint Ṭalḥa to marry her.58 ʿĀʾisha bint Ṭalḥa appears elsewhere together with Ḥubbā, as we have seen. These early Umayyad personalities were often used as characters in humorous stories and statements such as this one should probably not be read as truth. The claim that Ḥubbā was called Eve is possibly taken from Maḥāsin, a thoroughly unreliable book. Concluding remarks The historical reports and the anecdotes provide us with a timeline of the life of Ḥubbā. She was married at the time of ʿUthmān ibn ʿAffān (r. 644–656), when she accompanies the caliph on his pilgrimage from Medina to Mecca. About twenty years later, in 674, she is arguing with the poet Hudba ibn Ḫašram in Medina, who accuses her for having been attracted to Ibn ʾUmm Kilāb. According to Jāḥiẓ, she was almost fifty years old when she married this man. In 689, she meets with the caliph ʿAbd al-Malik. She is now a well-known socialite and the young aristocratic men of Medina use to gather in her house. She meets with the caliph again in 692, around forty years after the pilgrimage with ʿUṯmān. This powerful woman, who dared to defy the caliph, was defamed by Ṣāliḥ ibn Ḥassān, who was himself born in Medina under the Umayyads, but made his fortune in early ʿAbbāsid Kufa. His reports about her shamelessness are elaborated upon in classical Arabic literature. She was shameless already in the stories recounted by Jāḥiẓ and Ibn ʾAbī Ṭāhir, but at least she was sitting indoors talking to a select audience. In Nihāyat al-ʾarab by al-Nuwayrī (d. 733/ 1333), she walks out in the middle of the night and meets someone who asks

57 Pellat, “Ibn Harma”. 58 ʾAbū l-Faraj al-ʾIṣfahānī, ʾAġānī, vol. 11, 127.

Who was Ḥubbā al-Madīniyya?

341

why she is out that late. She answers, “I don’t care. If I meet a šayṭān, I will obey him, or if I meet a man, I will crave him.”59 Already in the late tenth century, the erotic Ḥubbā was elaborated upon through explicit sexual details – if, indeed, Jawāmīʿ al-laḏḏa was written then. For Ibn Naṣr, Ḥubbā is one of the old Arab authorities, and his representation of her is entirely positive – in line with his positive attitude toward sexuality, not least female sexuality. For him, it might have been a sign of civilization that this Ḥubbā taught her daughter and her son-in-law how to please each other and thus secure their marital love. She is less positively depicted by alMaydānī, who is the main source of information for modern scholars. The claims al-Maydānī makes in his notice on Ḥubbā, however, are not particularly trustworthy. The claim that she was called Eve may be a misreading, and the claim that she taught the women different types of coition is a later interpretation of al-Jāḥiẓ’s words for groaning and moving. It is remarkable that Fedwa Malti-Douglas considers her a “prototypical character” embracing all women in adab on the basis of the twelfth-century al-Maydānī’s account. The much earlier Ibn ʾAbī Tāhir, whom she also refers to, might have been countered with representations of Ḥubbā by his contemporaries, al-Balāḏurī and Ibn Qutayba, or the somewhat later al-Ṭabarī. Ibn ʾAbī Ṭāhir himself quotes Ḥubbā’s wisdom, which Malti-Douglas notes, which should have made her consider that the obscenities attributed to Ḥubbā have more to do with genre than with her person. Yet, the creation of Ḥubbā’s erotic persona is problematic, and I concur with Malti-Douglas’s line of argument to a certain degree. Why was she assigned those obscene scenarios, such as having sex with a dog, if she was, as alBalāḏurī claims, a society lady and type of spokeswoman for the fityān, the young men of Quraysh? The wisdom attributed to her indicates that she was considered a judicious and knowledgeable woman, which may be the cause for the fityān to visit her; yet, later authors of collections of proverbs maintain that her fame comes from being horny. There seem to be at least some degree of misogyny operating here; perhaps she was defamed because she was a social woman with younger husband. Jāḥiẓ asserts that Ḥubbā married a man who was younger than she, and later authors make the age difference even bigger. However, the poem that suggests Ḥubbā loved Ibn ʾUmm Kilāb for his youth and strong arms is a lampoon

59 Al-Nuwayrī, Nihāyta, vol. 4, 22.

342

Pernilla Myrne

and nothing guarantees that it is true. We do not know for certain the name of Ḥubbā’s husband, but Jāḥiẓ claims that Ibn ʾUmm Kilāb is another nasab for Ibn Lisān al-Ḥummara, a man with a degree of fame. He was renown in his time for wisdom and eloquence. Like Ḥubbā, he became proverbial and is mentioned by al-Maydānī, but for his profound knowledge of genealogy.60 He was also known for his “gift for lively repartee”, like Ḥubba.61 He never made it into the famous history books; however, in contrast to Ḥubbā, whose persona might have been easier prey for storytellers.

References Primary sources al-ʿAskarī, ʾAbū Hilāl, Jamharat al-ʾamṯāl. Edited by Muḥammad ʾAbū l-Faḍl ʾIbrāhīm and ʿAbd al-Majīd Qaṭāmish, 2 vols., Cairo: al-Muʾassasa al-ʿarabiyya al-ḥadīṯa lil-ṭabʿ wa-l-našr, 1384/1964. Ibn ʾAbī Ṭāhir al-Ṭayfūr, Balaġāt al-nisāʾ. Edited by Barakāt Yūsuf Habbūd. Beirut: AlMaktaba al-ʿAṣriyya, 2005. al-Balāḏurī, ʾAnsā al-ʾašrāf. Volume V. Edited by Iḥsān ʿAbbās Beirut: Dār al-Našr, 1996. Vol. XI. Edited by W. Ahlwardt under the title Anonyme arabische Chronik. Greifswald: Selbstverlag, 1883. al-Buḫārī, Muḥammad ibn ʾIsmaʿīl, Ṣaḥīḥ al-Buḫārī, vol. 7. Edited by Muṣṭafā al-Buġā, Damascus: Dār Ibn Kaṯīr, 1410/1990. Ibn Manẓūr, Muḥammad ibn Mukarram, Lisān al-ʿarab, vol. 5; first edition Cairo, 1883– 1891, edition used published in Beirut: Dār Ṣādir, n.d. Ibn Naṣr al-Kātib, ʿAlī, Jawāmiʿ al-laḏḏa, Mss Aya Sofya, 3836, 3837, and Fatih, 3729. Ibn Qutayba, ʿAbdullāh ibn Muslim, ʿUyūn al-ʾaḫbār. Edited by Yūsuf ʿAlī Ṭawīl, 4 parts in 2 vols., Beirut: Dār al-kutub al-ʿilmiyya, 1985. Ibn Saʿd, Muḥammad, Kitāb al-Ṭabaqāt al-kabīr. Edited by Eduard Sachau, 9 vols., Leiden: Brill, 1904–1940. al-ʾIṣfahānī, ʾAbū l-Faraj, al-Kitāb al-ʾAġāni. Edited by ʾIḥsān ʿAbbās et al. 25 vols., Beirut: Dār Ṣādir, 1423/2002. al-ʾIṣfahānī, Hamza, al-Durra al-fāḫira fī l-ʾamṯāl al-sāʾira. Edited by ʿAbd al-Majīd Qaṭāmish, 2 vols., Cairo: Dār al-Maʿārif, 1971-1972. Jāḥiẓ, ʾAbū ʿUṯmān ibn Baḥr, Ḥayawān. Edited by ʿAbd al-Salām Hārūn, 7 vols., 2nd ed. Cairo: Muṣtafā al-Bābī al-Ḥalabī, 1966.

60 Hamza al-Iṣfahānī, Durra, 391. Al-Maydānī, Majmaʾ, vol. 3, 396. 61 Pellat, ”Ibn Lisān al-Ḥummara”.

Who was Ḥubbā al-Madīniyya?

343

Jāḥiẓ, ʾAbū ʿUṯmān ibn Baḥr. “Al-Mufāḫara” In Rasāʾil al-Jāḥiẓ. Edited by ʿAbd alSalām Muḥammad Hārūn. 3 vols., Cairo: Maktabat al-Ḫānjī, 1399/1979. Ps.-Jāḥiz, Kitāb al-maḥāsin wa-l-ʾaḍdād. Edited by Gerlof van Vloten; Leiden: Brill, 1898. al-Mubarrad, Muḥammad ibn Yazīd. Fāḍil. Edited by ʿAbd al-ʿAzīz al-Maymanī, Cairo: Dār al-Kutub, 1956. al-Maydānī, ʾAḥmad ibn Muḥammad, Majmaʿ al-amṯāl. Edited by Muḥammad ʾAbū lFaḍl Ibrāhīm, 4 vols., Cairo: al-Bābī al-Ḥalabī, 1959. al-Nuwayrī, ʾAḥmad ibn ʿAbd al-Wahhāb. Nihāyat al-ʿarab fī funūn al-ʾadab, n.e., 33 vols., Cairo: Dār al-kutub, 1923-98. al-Ṭabarī, The History of al-Ṭabarī, Vol. 21, The victory of the Marwānids. Translated by Michael Fishbein. New York: State University of New York Press, 1990.

Secondary sources al-ʾAlbānī, Muḥammad Nāṣir al-Dīn. The Salah in the Light of the Prophet’s Tradition. Translated by Usama ibn Suhaib Hasan, Islamic Book Trust Kuala Lumpur, 2004. Borg, Gert. “Lust and Carnal Desire: Obscenities attributed to Arab women”, in Arabic and Middle Eastern Literatures, Vol. 3, No. 2, 2000. Gériès, Ibrahim. “al-Bayhaqī, ʾIbrāhīm b. Muḥammad”, Encyclopaedia of Islam, 3rd edition. Edited by Gudrun Krämer et. al., Leiden: Brill Online, 2014. Gériès, Ibrahim. Un genre littéraire arabe: al-Mahāsin wa-l-masāwī. Paris: G.-P. Maisonneuve et Larose, 1977. Leder, Stefan. Das Korpus al-Haiṯam ibn ʾAdī (st. 207/822): Herkunft, Überlieferung, Gestalt früher Texte der ahbār Literatur. Frankfurt am Main: Klostermann, 1991. Malti-Douglas, Fedwa. Gender and Discourse in Arabo-Islamic Writing, Cairo: The American University in Cairo Press, 1992 (first published by Princeton University Press. 1991). al-Munajjid, Ṣalāḥ al-Dīn. 1958. Al-ḥayāt al-jinsiyya ʿinda l-ʿarab. Beirut: Dār al-Kutub. Pellat, Charles. ”Al-Hayṯam b. ʿAdī” in Encyclopaedia of Islam, 2nd edition, vol. 3. Edited by Bernard Lewis et al., 328. Leiden: Brill. Pellat, Charles. ”Djins”, Encyclopaedia of Islam, 2nd edition, vol. 2, 551. Leiden: Brill. Pellat, Charles. ”Ibn Lisān al-Ḥummara”, Encyclopaedia of Islam, 2nd ed, Vol. 3. Edited by Bernard Lewis, et al., 854. Leiden: Brill. Pellat, Charles. “Ibn Harma”. Encyclopaedia of Islam, 2nd ed, Vol., 3. Edited by B. Lewis, et al., 786. Leiden: Brill. Pellat, Charles. “Nouvel essai d’inventaire de l’oeuvre Gahizienne”, Arabica, vol. 31, No. 2, 117–164, 1984. Rosenthal, Franz. ”Male and Female: Described and Compared.” In J. W. Wright Jr. and Everett Rowson, Homoeroticism in Classical Arabic Literature. New York: Columbia University Press, 1997. Rosenthal, Franz. ”Ḥamza al-Iṣfahānī”. Encyclopaedia of Islam, 2nd ed., vol., 3. Edited by Bernard Lewis et al., 156. Leiden: Brill.

344

Pernilla Myrne

Rowson, Everett K. “Arabic: Middle Ages to Nineteenth Century”, in Gaétan Brulotte and John Phillips (eds.), Encyclopedia of Erotic Literature vol. I, New York: Routledge, 2006, 43–61. Scott Meisami, Julie. “Writing Medieval Women.” In Julia Bray (ed.), Writing and Representation in Medieval Islam. Muslim horizons. London and New York: Routledge, 2006. Sezgin, Fuat. Geschichte des arabischen Schrifttums. 13 vols., Leiden: Brill, 1967–1984.

Überlegungen zur Erinnerung und Identitätsbildung im Koran Georges Tamer, Universität Erlangen-Nürnberg

Historische Religionen zehren aus Erinnerungen. Bei einer jeden von ihnen steht am Anfang ein übernatürliches Ereignis, das den Beginn einer Tradition markiert; durch dessen konstante Erinnerung wird eine jeweils unverwechselbare Identität konstituiert. Da der Gegenstand der folgenden Überlegungen, die Religion des Islams, zu den monotheistischen Traditionen gehört, beschränke ich mich in diesen einleitenden Bemerkungen auf diese Traditionen, um danach die Stellung der Erinnerung und deren identitätsstiftende Bedeutung im Koran zu beleuchten. *** So bedeutend für das Bild Gottes und den Sinn der Welt biblische Geschichten von universeller Dimension wie der Schöpfungsbericht und die Sintflut-Erzählung sind, wird die Vorstellung vom Volk Gottes erst durch gemeinschaftsspezifische Ereignisse allmählich gebildet, deren Erinnerung im Mittelpunkt des Judentums steht. Bemerkenswert ist es, dass die Initiierung solcher Ereignisse auf eine ursprüngliche Offenbarung Gottes zurückgeht: Er offenbart sich Abram, schließt mit ihm einen Bund und verheißt ihm einen Sohn, der mit seinen Nachkommen an der Verwirklichung der Verheißungen Gottes Anteil nehmen soll. 1 Eine Nachricht, die die betagte Sara zum Lachen bringt. Daran erinnert der Name, yiṣḥāq, wahrscheinlich aus yiṣḥāq-ʾēl, im Sinne von „El hat jemanden zum Lachen gebracht.“2 Mit dem Gebot der Beschneidung wird den männlichen Angehörigen des Bundes die Erinnerung daran am Leib festgehalten. An den Bund mit Abraham und den Patriarchen erinnert sich Gott und rettet sein Volk aus der Knechtschaft, in die es geraten ist, weil eben diese Erinnerung in seinem Bewusstsein oft nicht mehr präsent war.3 Erinnerung gleicht in                                                                                                                 1

Das Buch Genesis, Kap. 12-13.

2

Andreas Michel, „Isaak“, in: Michaela Bauks, Klaus Koenen, Stefan Alkier (Hg.), Das wissenschaftliche Bibellexikon im Internet (WiBiLex), Stuttgart 2006 ff.

3

Davon handeln vor allem die späten Propheten Israels. Siehe beispielsweise: Jesaja 48:1-16; 50:1-3; 65:1-7.

346

Georges Tamer

diesem Zusammenhang Treue. Pascha, das Fest der Erinnerung an den Übergang von der Knechtschaft in die Freiheit, und das Gedenken an die Begegnung zwischen Mose und dem Herrn auf dem Berg Sinai, bei der der eifrige Prophet die zehn Gebote auf zwei Tafeln erhielt, bilden zentrale Geschehnisse der religiösen Geschichte des Judentums.4 Dessen beanspruchte Sonderstellung inmitten heidnischer Völker kommt durch die zyklische Erinnerung daran zum Ausdruck. Jeden Schabbat entsinnen sich Juden der ihnen zuteil gewordenen Gnade Gottes im Bewusstsein ihrer eigenen Identität. Konstitutives Erinnern an entscheidende Ereignisse der Heilsgeschichte wird auch im Christentum um weitere Dimensionen erweitert gepflegt. An Lehren aus der Vergangenheit zu erinnern und sie mit neuem Geist zu erfüllen hat Jesus von Nazareth schon in der Bergpredigt unter Beweis gestellt.5 Mit dem Fest seiner Auferstehung erhält das Pascha-Fest die tiefgreifende Bedeutung des Übergangs von der Knechtschaft der Sünde zur Freiheit der Kinder Gottes.6 Die Erinnerung daran ist mit der Erinnerung an das Kreuz unzertrennlich verbunden. Beide Ereignisse bilden zwei Seiten eines unverkennbaren Merkmals des Christentums, das damit die Verheißung in die Welt setzt, dass die Realität des Todes von der Realität seiner Überwindung durchdrungen ist. Ein Paradoxon, das in Glaube und Hoffnung angenommen wird und in aufopfernden Taten der Nächstenliebe seinen Ausdruck findet. Eine christlich aufgefasste Erinnerung (Anamnese) ist mystische Vergegenwärtigung; das Heilsgeschehen Christi wird in der Eucharistie wiedergeholt; es wird in die gegenwärtige Existenz des Gläubigen und der Gemeinde eingepflanzt, sodass das, was einst war, hier und jetzt wie damals real und lebendig wird. Jeden Sonntag feiern Christen die Auferstehung; daran ist die Erinnerung gekoppelt, dass sie durch ihre Taufe dazu berufen sind, die Freude an der Überwindung des Todes selbst zu erleben und sie anderen weiter zu vermitteln. In der byzantinischorthodoxen Liturgie wird am Sonntag verkündet, dass „heute“, d.h. an diesem und jedem Sonntag, an dem die Auferstehung gefeiert wird, „der Welt die Erlösung“ real zuteil geworden ist. Eine gelebte Erinnerung also, die die Identität des ernst genommenen Christseins kennzeichnet.                                                                                                                 4

Rela M. Geffen, „Passover“, in: Michael Berenbaum/Fred Skolnik, Encyclopaedia Judaica², Detroit: Macmillan Reference USA, 2007, S. 678-683.

5

Das Evangelium nach Matthäus 5: 21-46.

6

1. Brief des Apostels Paulus an die Korinther 5:7f.

 

Überlegungen zur Erinnerung und Identitätsbildung im Koran

347

Der Islam ist als dritte Offenbarungsreligion Teil derselben religiösen Tradition, die auch das Judentum und das Christentum umfasst. In diesem Sinne könnte der Koran, nach der Thora und dem Evangelium, als das dritte Testament des einen und desselben Bundes bezeichnet werden, den Gott durch Abraham und weitere Gestalten der jüdisch-christlichen Heilsgeschichte mit den Menschen geschlossen hat. Im Koran werden große Teile dieser Geschichte eigenartig in Erinnerung gerufen, vereinnahmt oder revidiert. An dessen Umgang mit solchen Erinnerungen wird der Prozess der Identitätsbildung in der Anfangsphase des Islams reflektiert. Die Anfänge der Identitätsbildung im Islam verliefen in verschiedenen Etappen vor allem in abgrenzender Auseinandersetzung mit dem religiösen und kulturellen Umfeld, in dem der Islam entstanden war. Für die Entwicklung einer religiös-islamischen Identität war die Ablehnung des Heidentums maßgeblich ebenso wie die allmählich vollzogene Abgrenzung gegenüber Judentum und Christentum. Dieser historischen Tatsache trägt das koranische Gedächtnis Rechnung, indem sich der Koran mit zwei Traditionslinien auseinandersetzt: einer gottgläubigen – jüdischen, christlichen und altarabischen – und einer polytheistischen heidnischen. Heidnische Überzeugungen werden unmittelbar nach ihrer Erwähnung widerlegt, wie es beispielsweise mit der vorislamisch-arabischen Auffassung, die schicksalhafte endlose Zeit, dahr, richte die Menschen zugrunde, geschieht:7 Sie sagten: „Es gibt nichts anderes als unser Leben hier in dieser Welt. Wir sterben, und wir leben. Es ist allein die Zeit, die uns zugrunde richtet.“ Doch haben sie kein Wissen darüber, sie gehen allein Vermutungen nach. Wenn ihnen unsere Verse als Beweis vorgetragen werden, dann ist ihr einziges Argument: „Bringt unsere Väter herbei, wenn ihr die Wahrheit sagt.“ Sprich: „Gott gibt euch das Leben, dann lässt er euch sterben, dann versammelt er euch zum Tag der Auferstehung, über den kein Zweifel herrscht.“ Die meisten Menschen jedoch haben kein Wissen. Die koranische Auseinandersetzung mit monotheistischen Traditionen gestaltet sich viel komplexer, wie wir im weiteren Verlauf dieses Beitrags sehen                                                                                                                 7

Q 45:24-26. Stellen aus dem Koran werden nach der Übersetzung von Hartmut Bobzin, Der Koran, München: C.H. Beck 2010 zitiert und wenn nötig modifiziert. Vgl. die Analyse der Stelle in Georges Tamer, Zeit und Gott. Hellenistische Zeitvorstellungen in der altarabischen Dichtung und im Koran, Berlin/New York: Walter de Gruyter 2008, S. 193-197.

348

Georges Tamer

werden. Aus vorislamisch-arabischen Gedichten wissen wir, dass die Araber im Entstehungskontext des Islams die Erinnerung an vergangene Personen und Ereignisse gepflegt haben. Zum Beginn seines berühmtesten Gedichts, almuʿallaqa, erinnert sich Imruʾ al-Qais, der Fürst aus dem christlichen Königreich Kinda im südlichen Teil der Arabischen Halbinsel des sechsten Jahrhunderts n. Chr. an seine inzwischen an einen anderen Ort gewanderte Geliebte und deren verödete Wohnstätte:8

‫ﻮﻝل ﻓَ َﺤﻮْ َﻣ ِﻞ‬ ِ ‫ﺑِ ِﺴ ْﻘ ِﻂ ﺍاﻟﻠﱢ َﻮﻯى ﺑَ ْﻴﯿﻦَ ﺍاﻟ ﱠﺪ ُﺧ‬ ‫ﺄﻝل‬ ٍ ‫ﻟِ َﻤﺎ ﻧَ َﺴ َﺠ ْﺘﻬﮭَﺎ ِﻣ ْﻦ َﺟﻨُﻮ‬ ِ ‫ﺏب ﻭو َﺷ ْﻤ‬ ‫ـﻞ‬ ِ ‫ﻳﯾَﻘُﻮْ ﻟُﻮْ ﻥنَ ﻻَ ﺗَ ْﻬﮭﻠِ ْﻚ ﺃأَ َﺳ ًﻰ َﻭوﺗَ َﺠ ﱠﻤ‬ ‫ﺱس ِﻣ ْﻦ ُﻣ َﻌ ﱠﻮ ِﻝل‬ ٍ ‫َﺍاﺭر‬ ِ ‫ﻓَﻬﮭَﻞْ ِﻋ ْﻨ َﺪ َﺭرﺳ ٍْﻢ ﺩد‬

‫ﺐ ﻭو َﻣ ْﻨ ِﺰ ِﻝل‬ ٍ ‫ْﻚ ِﻣ ْﻦ ِﺫذ ْﻛ َﺮﻯى َﺣﺒِﻴﯿ‬ ِ ‫ﻗﻔَﺎ ﻧَﺒ‬ ‫ﺿ َﺢ ﻓَﺎﻟ ِﻤ ْﻘﺮﺍا ِﺓة ﻟَ ْﻢ ﻳﯾَﻌْﻒُ َﺭر ْﺳ ُﻤﻬﮭﺎ‬ ِ ْ‫ﻓَﺘُﻮ‬ ‫ﱠ‬ ‫ﺻﺤْ ﺒِﻲ ﻋَﻠﻲ َﻣ ِﻄﻴﯿﱠﻬﮭُـ ُﻢ‬ َ ‫ُﻭوﻗُﻮْ ﻓﺎ ً ﺑِﻬﮭَﺎ‬ ٌ‫ﻭوﺇإِ ﱠﻥن ِﺷﻔـَﺎﺋِﻲ َﻋﺒْـ َﺮﺓةٌ ُﻣ ْﻬﮭ َﺮﺍاﻗَـﺔ‬

Die wehmütige Erinnerung, die den Dichter viele Tränen vergießen lässt, teilt er mit Reisegefährten, die er gleich im ersten Wort des Gedichts dazu auffordert, innezuhalten und mit ihm zu weinen. In dialogischer Erinnerung artikuliert sich das lyrische Ich, das die schmerzliche Trennung und zerstörerische Wirkung der Zeit mittels der Poesie zu bewältigen versucht. Dies war dem Propheten des Islam durchaus bekannt. Ihm selber wurde vorgeworfen, er sei einer dieser Dichter, die ihre Inspiration von Dämonen erhielten, was dazu geführt hätte, dass seine Offenbarungen für dämonisch gehalten worden wären.9 Der Koran weist solche Vorwürfe vehement zurück:10                                                                                                                 8 Abū ʿAbdillāh al-Ḥusain ibn Aḥmad az-Zauzanī, Šarḥ al-muʿallaqāt as-sabʿ, Beirut: Dār al-Bayān 1973, S. 7f., 10f. Übersetzung: Imruʾ al-Qais, Die Muʿallaqa, Übersetzung von Salomon Gandz. Wien: Hölder, 1913, S. 10-17: „Haltet an, ihr beiden, auf daß wir weinen in Erinnerung an eine Geliebte und Wohnstätte am Abhange des gekrümmten Sandhügels zwischen ad-Daḫūl und Ḥaumal, Und Tūḍiḥ und al-Miqrāt, noch unverwischt ist ihre Spur, wenn auch kreuz und quer über sie dahinfuhr Süd und Nord. Indem nun meine Genossen ihre Reittiere bei mir daselbst anhielten, sprachen sie: Vergeh‘ nicht vor Kummer und fasse dich in Geduld. Und wahrlich Heilung ist mir die vergossene Träne. Doch ist denn eine verwischte Wohnungsspur ein Ort, um sich dem übermäßigen Weinen hinzugeben?“ 9 Q 37:36: „Und sie sprachen: ‚Sollen wir wohl von unseren Göttern lassen zugunsten eines besessenen Dichters?“; Q 69:41: „Es ist nicht das Wort eines Dichters – wie wenig glaubt ihr doch!“

 

Überlegungen zur Erinnerung und Identitätsbildung im Koran

349

Wir lehrten ihn nicht die Dichtung, denn das gebührt auch ihm nicht. Siehe, es ist nichts anderes als eine Mahnung (ḏikr) und eine klare (bzw. klärende) Lesung (qurʾān mubīn). Er erinnert sich der Dichter in einer gleichnamigen Sure und wirft ihnen Verirrung und großes Gerede ohne Taten vor; ihnen folgen die Irrenden.11 Dennoch finden sich im Koran Motive, die in der altarabischen Dichtung geläufig waren. Ein Beispiel dafür ist das Ubi-sunt-Motiv, wonach durch die Erinnerung an vergangene Völker und Herrscher die allen Menschen gemeinsame Vergänglichkeit veranschaulicht wird.12 Daraus sollten nützliche Lektionen für den Umgang mit der Gegenwart gezogen werden. Im Koran fungiert die Erinnerung an vergangene Völker, die sich Gottes Propheten widersetzten und darum göttliche Strafen erleiden mussten, als Mahnung für die Gegner Muḥammads.13 Wenn von Erinnerung und Identitätsbildung im Islam gesprochen wird, verdient der Begriff islām besondere Aufmerksamkeit. Das vom arabischen Verb aslama im Sinne von „übergeben, unterwerfen, sich ergeben, sich hingeben“ abgeleitete Verbalsubstantiv islām bedeutet „Unterwerfung, Hingabe“. Mit diesem theologischen Konzept wird der Urzustand der Menschheit bezeichnet, die als Geschöpf von Beginn an ihrem Schöpfer gegenüber in ergebe                                                                                                                                                                                                                                                                                                                                     10 Q 36:69. 11 Q 26:224-226. Vgl. die interessante Debatte zwischen Michael Zwettler und Irfan Shahid zu Sure 26: Michael Zwettler, „A Mantic Manifesto. The Sūra of the Poets and the Qurʾānic Foundation of Prophet Authority“, in: James L. Kugel (ed.), Poetry and Prophecy: the Beginnings of a Literary Tradition, Ithaca: Cornell University Press, 1990, 75-119; Irfan Shahîd, „The Sūra of the Poets, Qurʾān XXVI: Final Conclusions“, in: Journal of Arabic Literature 35 (2004), 175-220; Michael Zwettler, „The Sūra of the Poets: Final Conclusions?“, in: Journal of Arabic Literature 38 (2007), 111-166; Irfan Shahîd, „The Sūra of the Poets Revisited“, in: Journal of Arabic Literature 39 (2008), 398-423. 12 Q 3: 137; 9: 69; 10: 13. Ausführlicher dazu Georges Tamer, Zeit und Gott. Hellenistische Zeitvorstellungen in der altarabischen Dichtung und im Koran, Berlin/New York: Walter de Gruyter 2008, S. 46-48. 13 Ein Beispiel ist Šuʿayb, der zu dem Volk Midian gesandt wurde Q 7:90-92: „Die Ältesten aus seinem Volk, die ungläubig waren, sprachen: ‚Wenn ihr nun wirklich Schuʿaib folgen wollt, siehe, dann seid ihr Verlierer!‘ Da erfasste sie das Beben, und am Morgen lagen die hingestreckt in ihrem Haus, die Schuʿaib einen Lügner nannten, als ob sie nie darin gewesen wären. Die Schuʿaib einen Lügner nannten, sie wurden die Verlierer!“

350

Georges Tamer

ner Haltung existiert.14 Der Mensch ist demnach von Natur aus derart veranlagt, dass er sich Gott hingibt. Das Wort islām bedeutet also ursprünglich nicht die institutionalisierte, von Muḥammad gegründete Religion im Arabien des 7. Jahrhunderts nach Christus, sondern einen natürlichen, Gott ergebenen Zustand des von Gott geschaffenen Menschen. Diese Bedeutung gilt auch für die koranische Bezeichnung vorislamischer Personen als Muslime. Wenn Noah oder Jesus und seine Jünger jeweils mit dem arabischen Partizip Präsens Aktiv muslim, Pl. muslimīn, bezeichnet werden, bedeutet dies nicht, dass sie Anhänger der Religionsgemeinschaft Muḥammads, sondern dass sie Gott ergebene Menschen waren.15 Die hier angedeutete Doppeldeutigkeit des Begriffs islām und zusammenhängender Termini verleitet zu einseitigen Deutungen, wofür begriffliche Ambiguitäten besonders geeignet sind.16 Aber genau diese Doppeldeutigkeit ist im gegenwärtigen Zusammenhang relevant. Muḥammads Botschaft zielte darauf ab, den ursprünglichen Zustand der absolut-monotheistischen Selbsthingabe des Menschen gegenüber Gott wiederherzustellen. Der von Muḥammad beschworene islām ist an sich eigentlich eine Erinnerung an die Urhaltung der Menschheit, die keinen anderen Gott außer den einen Gott, arabisch: Allāh,17 kennen und anbeten sollte, um diese ursprüngliche Haltung der Menschheit wiederzubeleben. Dieser Glaubensinhalt bildet tatsächlich den ersten Teil des islamischen Glaubensbekenntnisses, aššahāda: lā ilāha illā llāh, es gibt keinen Gott außer Gott. Schwingt im Wort islām die Erinnerung an den Urzustand der menschlichen Hingabe Gott gegenüber mit, korrespondiert dies damit, dass der Koran von nachdrücklicher Aufforderung zur Erinnerung und von umfangreichen Erinnerungen an vergangene                                                                                                                 14 Q 3:83. 15 Vgl. für Noah: „Falls ihr euch dann abwendet – Ich habe keinen Lohn von euch verlangt. Meine Belohnung obliegt nur Gott allein; und mir ward befohlen, ein Gottergebener zu sein (an akūna mina al-muslimīn).“ (Q 10:71). Vgl. für Jesus: „Und damals, als ich den Jüngern eingab: ‚Glaubt an mich und meinen Gesandten!‘ Da sprachen sie: ‚Wir glauben! Bezeuge, dass wir gottergeben (muslimūn) sind!‘“ (Q 5:111). 16 Vgl. dazu Muhammad Arkoun: „Islam“, in: Jane D. McAuliffe (Hg.), Encyclopaedia of the Qurʾān, Bd. 2. Brill: Leiden 2005, S. 565-570, mit weiterführender Literatur. 17 Vgl. Aziz Al-Azmeh, „Linguistic Observations on the Theonym Allāh“, in: Bilal Orfali (Hg.): In the Shadow of Arabic. The Centrality of Language to Arabic Culture, Leiden and Boston: Brill 2011, S. 267-281.

 

Überlegungen zur Erinnerung und Identitätsbildung im Koran

351

Heilsgeschehen geprägt ist. In diesem Sinne lässt sich der Koran zu Recht als Erinnerungsstimulator bezeichnen.18 Die geistige Tätigkeit der Erinnerung drücken im Arabischen Verben und Substantive aus, die von der Wurzel ḏ-k-r abgeleitet werden (z.B. ḏikr, ḏikrā, taḏkira, uḏkur, uḏkurū usw.). Sie kommen in insgesamt 274 Versen des Korans vor. Hinzu kommen weitere Ausdrücke und Redewendungen, die die Erinnerung implizieren oder dazu auffordern.19 Darunter sind zahlreiche Verse, die mit dem Adverb iḏ eingeleitet sind. Sie rufen vergangene Personen und Geschehnisse in Erinnerung.20 An einigen der frühesten Stellen wird der Koran in selbstreferentiellen Aussagen als Erinnerung bezeichnet, wie es in Sure 74 alMuddaṯṯir, die traditionell mit dem Beginn der aktiven Mission Muḥammads in Zusammenhang gebracht wird, vorkommt:21 O nein! Siehe, das ist eine Mahnung (taḏkiratun), wer will, der ruft sie in Erinnerung (ḏakarahu). Nur, wenn Gott es will, denken sie daran. Er ist es, den man fürchten muss, und er ist fähig zu vergeben. In einer weiteren frühen Sure, nämlich Sure 81 at-Takwīr, in der nach gewaltigen eschatologischen Bildern von einer Muḥammad zuteilgewordenen Engelsvision knapp berichtet wird, wird der Koran als eine „mahnende Erinnerung“ (ḏikr), die an die Bewohner der Welt gerichtet wird, bezeichnet.22 Darüber, ob sie die Mahnung wahrnehmen, entscheiden die Menschen jedoch freiwillig.23 Wie es sich ziemt, steht Gott im Mittelpunkt der Erinnerung im Koran und dies geschieht im Zusammenhang mit der Rezitation der Offenbarungen und                                                                                                                 18 Angelika Neuwirth, „Qurʾān, Crisis and Memory. The Qurʾānic path towards canonization as reflected in the anthropogonic accounts“, in: Angelika Neuwirth und Andreas Pflitsch (Hg.), Crisis and Memory in Islamic Societies. Proceedings of the third Summer Academy of the Working Group Modernity and Islam held at the Orient Institute of the German Oriental Society in Beirut. Würzburg: Ergon Verlag, 2001, S. 113-152. 19 Michael A. Sells, „Memory“, in: Jane D. McAuliffe (Hg.), Encyclopaedia of the Qurʾān, Bd. 3, S. 272-274. 20 Siehe beispielsweise Q 2:30, 34, 49-51, 53-55, 58, 60f. passim. 21 Q 74:54-56. 22 Q 81:27. 23 Q 74:55: „Wer will, der ruft sie in Erinnerung.“ Q 80:12: Und wer will, gedenket seiner.

352

Georges Tamer

der Verrichtung des Gebets.24 Sich Gottes zu erinnern ist ein Merkmal des intakten Menschseins, das im Naturzustand in einem kommunikativen Zusammenhang mit Gott steht. Eine solche Erinnerung gewährleistet, dass auch Gott sich des Menschen erinnert:25 Gedenket nun meiner (fa-ḏkurūnī), so will ich euer auch gedenken (aḏkurkum)! Und dankt mir, und verleugnet mich nicht! Es ist in diesem Zusammenhang von Bedeutung, dass nach koranischer Auffassung die erste Sünde des Menschen darin bestand, aufgrund satanischer Versuchung Gott und Seine Gebote vergessen zu haben:26 Wir hatten früher schon mit Adam einen Bund geschlossen (ʿahidnā), doch er vergaß (fa-nasiya); wir fanden bei ihm keinen festen Willen. Satan ist im Koran die Ursache des Vergessens. Er trägt die Schuld, auch wenn der junge Diener von Moses den Fisch vergisst, den sie auf einer legendären Reise verspeisen sollen.27 Die koranische Auffassung, dass die erste Sünde Adams Folge des vom Satan verursachten Vergessens ist, impliziert, dass sich der Mensch nicht ungehorsam gegen Gott auflehnte, wie die biblische Geschichte vom Sündenfall vermittelt. Im Gegenteil: Adam unterlag im Koran der Versuchung, indem er den ursprünglichen Bund mit Gott vergaß. Dieser Bund besteht eigentlich in der mit dem Konzept islām beschriebenen Grundhaltung der Hingabe, die die Menschheit Gott im Naturzustand erweist. In diesem Zusammenhang sei darauf hingewiesen, dass das koranische Wort für Unglaube kufr, der Ungläubige ist kāfir, ursprünglich so viel wie „etwas bedecken, etwas unsichtbar machen“ bedeutet.28 Kāfir ist also derjenige, der die Erinnerung an Gott und sein Wirken in der Welt durch den Schleier des                                                                                                                 24 Q 29:45: „Trag vor, was dir aus dem Buch eingegeben wurde, und verrichte das Gebet! Siehe, das Gebet hält vom Schändlichen und vom Verwerflichen ab. Doch das Gedenken Gottes ist wahrlich noch bedeutender. Gott weiß, was ihr tut.“ 25 Q 2:152. 26 Q 20:115. Bobzins Übersetzung modifiziert. 27 Q 18:63: „Er sprach: ‚Hast du gesehen? Als wir unsere Zuflucht bei dem Felsen suchten, vergaß ich doch den Fisch! Nur Satan hat mich ihn vergessen lassen, dass ich mich an ihn erinnerte. Da nahm er seinen Weg ins Meer auf wundersame Weise.‘“ Vgl. zur Stelle: Brannon M. Wheeler, Moses in the Quran and Islamic Exegesis, London: Routledge Curzon 2002, S. 10-19. 28 Al-Ḫalīl b. Aḥmad al-Farāhīdī, Kitāb al-ʿAyn, hrsg. Mahdī l-Maḫzūmī und Ibrāhīm as-Sāmarrāʾī, 8 Bde., Beirut o. J., Bd. 5, S. 356f.

 

Überlegungen zur Erinnerung und Identitätsbildung im Koran

353

Vergessens verschwinden lässt. Seiner Selbstbezeichnung als Erinnerung entsprechend wirkt der Koran nachhaltig dagegen. Zahlreiche Verse erinnern an die Wunder der Natur wie etwa den Wechsel der Zeiten,29 die hohen Berge,30 die Bäume und ihre Früchte, 31 als Ausdruck der Gnade Gottes, die jeder Mensch persönlich erfahren kann. Als Stimme der Erinnerung wird der Koran nicht müde, gegen die menschliche Neigung zur Vergesslichkeit zu predigen. Diejenigen, die die Erinnerung und den Glauben bewahren, werden als vernünftig bezeichnet:32 Er ist es, der auf dich das Buch herabgesandt hat. Einige seiner Verse sind klar zu deuten - sie sind der Kern es Buches, andere sind mehrfach deutbar. Doch die, in deren Herzen Verirrung ist, die folgen dem, was darin mehrfach deutbar ist, um Zweifel zu erwecken und um es auszudeuten. Doch nur Gott kennt dessen Deutung. Und die im Wissen fest gegründet sind, die sagen: „Wir glauben daran. Alles kommt von unserem Herrn.“ Doch nur die Einsichtsvollen lassen sich ermahnen. Ein wichtiges Merkmal der koranischen Erinnerung an die Werke Gottes, vor allem die Schöpfung, besteht in den frühen Suren darin, dass die Erinnerung an eine solche kosmische Vergangenheit unmittelbar mit der Antizipation einer künftigen, mit der Bestrafung der Gegner Muḥammads beladenen Eschatologie verklammert wird.33 Die erinnerte Vergangenheit, die an den sichtbaren Teilen der Schöpfung festgemacht werden kann, fungiert in diesem Zusammenhang als eine gewisse Bürgschaft dafür, dass die als Drohung aufgestellte, dennoch unsichtbare Zukunft zu gegebener Zeit einbrechen wird. Unter dem Motto: Wie Gott es anfangs tat, wird Er es auch am Ende tun. Durch die rhetorisch kraftvolle Darstellung der Höllenstrafen sollten die Gegner Muḥammads abgeschreckt und zum Glauben an die Wahrheit seiner Mission bewogen werden.34 Dass die Angst vor dem göttlichen Gericht eigentlich den Glauben an Gottes schöpferische Tätigkeit voraussetzt, und dies den mekkanischen Heiden offenbar fehlte, wird im Koran dadurch ausgeglichen, dass die Wunder der                                                                                                                 29 Q 2:164; 10:6; 16:12; 17:12; 45:5. 30 Q 16:81; 27:88; 78:7; 79:32; 88:19. 31 Q 2:22; 16:67; 27:60; 36:80. 32 Q 3:7. Bobzins Übersetzung wurde modifiziert. 33 Siehe beispielsweise Q 80:17-42; 86:5-17; 90:3-20. 34 Q 30:9; 40:82; 47:10.

354

Georges Tamer

Natur quasi als offenes Buch der Werke Gottes jedem Menschen mittels sinnlicher Wahrnehmung zugänglich sind.35 Die Gleichsetzung von Vergessenheit und Sünde spielt auch eine Rolle in der Auseinandersetzung mit den Juden und den Christen, die in den späteren Suren allmählich an polemischer Schärfe gewinnt. In diesem Sinn werden die Juden, banū Isrāʾīl, die Kinder Israels, genannt, oft an die Wohltaten Gottes in alten Zeiten erinnert.36 Ein Höhepunkt der Polemik gegen sie kommt zum Ausdruck, wenn ihnen vorgeworfen wird, den Bund mit Gott gebrochen und die Worte ihrer heiligen Schrift verändert zu haben:37 Wegen ihres Bundesbruches haben wir sie verflucht und ihre Herzen verhärtet. Sie rücken Wörter weg von ihrem Platz und haben einen Teil von dem vergessen (nasū), womit sie ermahnt wurden (ḏukkirū bihi). Noch immer kannst du Verrat unter ihnen sehen, mit Ausnahme weniger von ihnen. So verzeih ihnen, und vergib! Siehe, Gott liebt diejenigen, die Gutes tun. Ihnen und den Christen ist gemeinsam, dass sie alle „einen Teil von dem vergessen haben, woran sie in Ermahnung erinnert worden waren“.38 Ein solches Vergessen spiegelt sich darin wider, dass sie sich Muḥammad nicht angeschlossen haben. Daran wird die Auffassung angeknüpft, dass die Verkündigung Muḥammads, der Koran, eine authentische Linie der religiösen Erinnerung bewahrt, die in den beiden älteren Religionsgemeinschaften verloren gegangen sei und die die spezifische Identität des Islams als rein monotheistischer Religion ausmache.39 *** Wichtiger Gegenstand der religiösen Erinnerung im Koran sind Geschichten von biblischen Figuren, an die in den verschiedenen Kontexten der Mission

                                                                                                                35 Q 2:164; 3:190. 36 Q 2:40: „Ihr Kinder Israel! Gedenket meiner Gnade, die ich euch erwies! Haltet ihr den Bund mit mir, so halte ich den Bund mit euch! Mich allein, mich fürchtet!“ Q 2:47: „Ihr Kinder Israel! Gedenket meiner Gnade, die ich euch erwies, und dessen, dass ich euch erwählte vor den Weltbewohnern!“ Q 2:122: „Ihr Kinder Israel! Gedenket meiner Gnade, die ich euch erwies, und dessen, dass ich euch erwählt vor den Weltbewohnern!“ 37 Q 5:13. 38 Q 5:13-14. 39 Q 3:3; 5:15.

 

Überlegungen zur Erinnerung und Identitätsbildung im Koran

355

Muḥammads mit unterschiedlicher Akzentuierung erinnert wird.40 An zahlreichen Stellen im Koran wird der Prophet Muḥammad ausdrücklich dazu aufgefordert, sich früherer Gestalten der monotheistischen Religionsgeschichte zu erinnern.41 Die Geschichten werden nicht systematisch in historischer Reihenfolge, sondern verstreut und mehrfach wiederholt erzählt. Variieren Einzelheiten bei wiederholter Erzählung, bleibt dabei die Art der Erinnerung im Wesentlichen dieselbe: Erzählte Ereignisse werden aus der Vergangenheit in die reale Gegenwart des koranischen Geschehens übertragen, damit sie in historischen Situationen des Wirkens Muḥammads bestimmte Funktionen erfüllen. Indem sie entfaltet werden, bilden sie Bausteine der sich im Aufbau befindlichen Identität der neuen Religion. Für diese Geschichten gilt allgemein, dass sie jeweils stückweise und in Schüben erzählt werden, wobei die früheste Version einen Nukleus bildet, um den herum neue Schichten allmählich wachsen, die weitere Einzelheiten enthalten, welche veränderten Umständen der koranischen Verkündigung entsprechen.42 Erinnerung geht also mit Identitätsbildung im Koran einher. Wie dies geschieht, ist anhand einer Darstellung der koranischen Geschichte Abrahams zu demonstrieren. Abraham ist eine zentrale Gestalt im Koran. An vielen seiner Lebensstationen können Stationen aus dem Leben Muḥammads abgelesen werden.43 Mit                                                                                                                 40 Vgl. dazu: Heinrich Speyer, Die biblischen Erzählungen im Qoran, Hildesheim/New York City/Zürich: Georg Olms Verlag 1988 [Nachdruck der Ausgabe von 1931 bei Gräfenhainichen]; Roberto Tottoli, Biblical Prophets in the Qur’ān and Muslim Literature, Richmond: Curzon, 2002. 41 Wie beispielsweise Q 19:16, 41, 51, 54: „Und gedenke im Buch der Maria: Da sie sich vor ihren Leuten an einen Ort im Osten zurückzog (…) Gedenke im Buch des Abraham! Siehe, er war gerecht und ein Prophet. (…) Gedenke im Buch des Mose! Siehe, er war erwählt und war Gesandter und Prophet. (…) Gedenke im Buch des Ismael! Siehe, er war der Verheißung treu und war Gesandter und Prophet.“ 42 Vgl. Angelika Neuwirth, „Vom Rezitationstext über die Liturgie zum Kanon. Zu Entstehung und Wiederauflösung der Surenkomposition im Verlauf der Entwicklung eines islamischen Kultus“, in: Stefan Wild (Hg.), The Qurʾan as Text, Leiden: Brill 1996, S. 69-105; Nicolai Sinai, „The Qurʾan as Process“, in Angelika Neuwirth, Nicolai Sinai, and Michael Marx (Hg.), The Qurʾān in Context, Leiden, Boston: Brill 2010, S. 407-440. 43 Abrahams Episoden werden von Nicolai Sinai, Fortschreibung und Auslegung. Studien zur frühen Koraninterpretation, Wiesbaden: Harrassowitz, 2009, S. 97-160, ausführlich analysiert, um die interpretative Entwicklung im Koran darzustellen.

356

Georges Tamer

Recht kann er „Spiegel des Propheten“ genannt werden.44 In einem aus seiner Biographie stammenden Spruch soll Muḥammad in Bezug auf Abraham gesagt haben: „Nie habe ich einen Mann gesehen, der mir ähnlicher war.“45 Wie kein anderer Prophet wird Abraham im Koran imām, d.h. Vorbild46, ḥanīf, d.h. jemand, der sich vom Götzendienst seines Volkes abtrennt und sich dem einen Gott wendet47 und ḫalīl Allāh, Freund Gottes48 genannt. Die früheste Erwähnung Abrahams im Koran bringt ihn in Verbindung mit Schriftrollen, die er hinterlassen haben soll.49 In Übereinstimmung mit ihrem Inhalt wird Muḥammad ausdrücklich aufgefordert, Gott zu preisen, und die Menschen daran zu erinnern.50 Wie es in den frühen Suren üblich ist, werden die eschatologische Strafe und Belohnung in den Mittelpunkt der Mahnung gestellt.51 Aus derselben frühen Periode der Verkündigung stammt ebenfalls die früheste koranische Erwähnung der Begegnung Abrahams mit seinen göttlichen Gästen. Sie kündigen ihm die Geburt eines Sohnes und die Vernichtung von Lots Volk an.52 In den Suren, die im späteren Verlauf des Wirkens Muḥammads in seiner Heimatstadt Mekka entstanden sind, konzentrieren sich die Abraham-Episoden auf seine Bekehrung vom Götzendienst, seine darauf folgenden Bemühungen, seinen Vater und sein Volk zum Glauben an den einen Gott zu bewegen und seine Entschlossenheit, die Nichtigkeit der Götzen zu demonstrieren, was zur Folge hatte, dass er sie zerschmetterte.53 In der ganzen Auseinandersetzung ist                                                                                                                 44 Martin Bauschke, Der Spiegel des Propheten, Frankfurt: Lembeck 2008. 45 Ebd. S. 16. 46 Q 2:124. 47 Q 2:135; Q 3:67; Q 16:120. 48 Q 4:125. 49 Q 87:9-19; 53:36-37. 50 Q 87:9-10: „So mahne – wenn die Mahnung nützt! Sich mahnen lässt, wer gottesfürchtig ist.“ 51 Q 87:11-17. 52 Q 51:24-37. 53 Q 19:41-48: „Gedenke im Buch des Abraham! Siehe, er war gerecht und ein Prophet. Damals, als er zu seinem Vater sprach: ‚Mein Vater! Warum verehrst du etwas, was nicht hört und sieht und dir auch nichts hilft? Mein Vater! Siehe, zu mir gelangte Wissen, das noch nicht zu dir kam! So folge mir, dass ich dich führe einen geraden Weg! Mein Vater! Diene nicht dem Satan! Siehe, der Satan widersetzt sich dem Erbarmer. Mein Vater! Ich fürchte, dass dich eine Strafe vom Erbarmer trifft und du ein Gefolgsmann des Satans wirst!‘ Er sprach: ‚Willst du meine Götter verlassen, Abraham? Wenn du nicht endlich aufhörst, werde ich dich steinigen! Doch

 

Überlegungen zur Erinnerung und Identitätsbildung im Koran

357

die Rede Abrahams homiletisch; er versucht sein Volk vom Götzendienst abzubringen, wie es auch Muḥammad tat. Eine weitere Gemeinsamkeit zwischen Abraham und dem Verkünder des Korans ist die Trennung vom eigenen Volk aufgrund des Glaubensunterschieds.54 An einer recht späten Stelle wirkt sich die Erinnerung an Abraham entlastend aus. Nach der Eroberung Mekkas entfachte sich unter den Gefährten Muḥammads ein Streit darüber, ob es rechtens sei, dass Muslime für Eltern und Verwandte beten, die vor ihrem Tod den Glauben an Gott nicht angenommen haben. Ein offenbartes Verbot setzte dem Streit ein Ende.55 Da auch Muḥammad das Grab seiner als Götzendienerin verstorbenen Mutter besucht haben soll, wird an Abrahams Vorbild erinnert, der für seinen als Feind Gottes bezeichneten Vater in Erfüllung eines Versprechens gebetet hatte.56 Im Koran begehen die Propheten keine Fehler. Ihre Erinnerung spielt eine konstruktive Rolle bei der Bildung einer rechtgläubigen Identität. Auch an die von Abraham beabsichtigte Opferung seines Sohnes wird im Koran erinnert – allerdings ohne den Namen des zu opfernden Sohnes zu erwähnen. Der Text liefert jedoch überzeugende Gründe dafür, dass es sich in Übereinstimmung mit der biblischen Vorlage um Isaak handelt.57 Warum wird also Isaaks Name verschwiegen, wenn im Koran an die großartige Glaubenstat Abrahams erinnert wird? In der Bibel steht Isaak an zweiter Stelle in der genealogischen Linie, die von Abraham über ihn und seinen Sohn Jakob zu den Vätern der zwölf Stämme Israels hinführt. Er ist Teil des heilsgeschichtlichen Erbes im Judentum und Christentum. In der patristischen Bibelexegese präfiguriert Isaaks vom Vater gewollte, jedoch durch göttliches Einwirken unvollende                                                                                                                                                                                                                                                                                                                                     nun meide mich für längere Zeit!‘ Er sprach: ‚Friede sei mit dir! Ich werde meinen Herrn für dich um Vergebung bitten! Siehe, er ist mir wohlgeneigt. Ich halte mich von euch getrennt und von dem, was ihr an Gottes statt anruft, und rufe einzig meinen Herrn an. Vielleicht werde ich dadurch, dass ich meinen Herrn anrufe, nicht unglücklich!‘“ Vgl. ebenfalls Q 6:74-83; 21:51-70; 26:69-89; 37:83-99. 54 Q 19:49. 55 Q 9:113. 56 Q 9:114. Theodor Nöldeke, Geschichte des Qurāns. Zweite Auflage bearbeitet von Friedrich Schwally. Erster Teil: Über den Ursprung des Qorāns. Leipzig: Dieterich 1909 (Nachdruck: Hildesheim etc.: Georg Olms Verlag 2000), S. 225. Siehe Tabaris Auslegung der Stelle. 57 Q 37:101-109. Vgl. Ayaz Afsar, „A Comparative Study of the Intended Sacrifice of Isaac/Ishmael in the Bible and the Qurʾān“, in: Islamic Studies 46 (2007): 483-498.

358

Georges Tamer

te Opferung die Opferung Jesu auf dem Kreuz durch seinen Vater.58 Dass der Name Isaaks aus der Erzählung der geplanten Opferung verschwindet, ist folglich als ein Schritt der Selbstabgrenzung des Islams dem Judentum und Christentum gegenüber zu deuten. Isaak ist zwar einer der im Islam verehrten Propheten, er besitzt jedoch keine Zentralstellung in der auf Abraham zurückgehenden Genealogie, mit der der Islam Anspruch auf das monotheistische Erbe Abrahams erheben kann. Isaak muss hier diese Stellung an seinen Bruder Ismael abgeben, der in verschiedenen Quellen als Stammvater der Araber, der nach ihm benannten Ismaeliten, bezeichnet wird.59 Er tritt nun an der Seite seines Vaters als Wahrer der monotheistischen Tradition auf. Gemeinsam mit seinem Sohn Ismael errichtet Abraham die Kaaba und stiftet den Pilgerritus – eine koranische Neuerung im Vergleich zur biblischen Geschichte.60 Immer wenn Muslime nach Mekka pilgern, ahmen sie das Beispiel der beiden nach. Darüber hinaus vollziehen die Pilger denselben Lauf, den Hagar, eine Nebenfrau Abrahams, zusammen mit dem kleinen Ismael fast verdurstet und verzweifelt zwischen den Hügeln aṣ-Ṣafā und al-Marwa hin und her auf der Suche nach Wasser vollzog. Sie trinken aus demselben Quellbrunnen Zamzam, den Gott damals auf wundersame Art entspringen ließ, damit Hagar und ihr Sohn trinken konnten. Eine kathartische Erinnerung also, die eng mit einer zentralen Komponente der islamischen Identität verbunden ist. In ihrem Stiftungsgebet des Pilgerritus bitten Abraham und Ismael Gott darum, sie beide zu Ergebenen (muslimaini) und aus ihrer Nachkommenschaft eine ergebene Gemeinschaft (umma muslima) zu machen. Sie bitten Gott weiterhin darum, aus der Mitte ihrer Nachkommen einen Gesandten entstehen zu lassen, der ihnen Seine Verse vorträgt, sie das Buch und die Weisheit lehrt und sie vom Götzendienst                                                                                                                 58 Siehe Johannes Chrysostomus, Des hl. Johannes Chrysostomus Homilien über die Genesis oder das erste Buch Mosis. Erster Band. Herausgegeben von Prinz Max, Herzog zu Sachsen. Paderborn: Druck und Verlag von Ferdinand Schöningh 1913, S. 696. Siehe ebenfalls Mark Sheridan (Hg.), Ancient Christian Commentary on Scripture. Old Testament II Genesis 12-50, Downers Grove: InterVarsity Press 2002, S. 110. 59 Genesis 25:12-18. In seinem monumentalen Werk The Arabs in Antiquity: Their History from the Assyrians to the Umayyads, London und New York: Routledge Curzon 2003, behandelt Jan Retsö Schilderungen der Araber im Alten Testament. Siehe besonders die Erörterung der Erzählungen über Ismael und seine Nachkommen, S. 220-228. 60 Q 2:124-127.

 

Überlegungen zur Erinnerung und Identitätsbildung im Koran

359

läutert61 – eine Bitte, der nach islamischer Auffassung mit der Sendung Muḥammads entsprochen wurde.62 Die von ihm gegründete Religionsgemeinschaft ist ein Abbild der Religionsgemeinschaft Abrahams, millat Ibrāhīm.63 In Mekka, wo die erste Phase der Verkündigung bis zur Auswanderung Muḥammads nach Medina verlief, war Ismael nicht wichtig. Erst infolge des Konflikts zwischen Muḥammad und den in Medina lebenden jüdischen Stämmen gewann er an Bedeutung. Dies geschieht gleichzeitig mit dem im Koran angeordneten Wechsel der Gebetsrichtung von Jerusalem nach Mekka, obwohl diese Stadt zu dieser Zeit noch heidnisch war.64 *** Zum Schluss kann Folgendes festgehalten werden: Das Konzept der Erinnerung im Koran ist zukunftsgerichtet. Darin evozierte Erinnerungen stellen sich als wichtige Bausteine einer sich allmählich kristallisierenden eigenen Identität in der Anfangsphase des Islams dar. Erinnerte Personen und Ereignisse erfüllen bestimmte Funktionen und fungieren als Leitbilder im koranischen Prozess der islamischen Identitätsbildung. Mit dem koranischen Begriff islām verbindet sich ohnehin ein theologisch-abstraktes Konzept der Erinnerung, mit der die ursprüngliche Identität des Menschseins im Verhältnis zu Gott wiederhergestellt werden soll, die im Verlauf der Geschichte des Monotheismus beschädigt worden war. Zu deren Wiederherstellung tragen im Koran ermahnende Erinnerungen bei. Der Koran operiert damit gegen das vom Satan verursachte Vergessen. Im Rekurs auf die vorislamische Geschichte des Monotheismus werden ebenfalls Erinnerungen evoziert, modifiziert und dem historischen Kontext entsprechend der koranischen Verkündigung eingesetzt. Ein konflikthaftes Unterfangen, denn es geht um das strittige Erbe Abrahams, des Patriarchen des Monotheismus. Der Rückgriff auf den Anfang bleibt im Islam auch gegenwärtig von erheblicher Relevanz. Ein Beispiel dafür ist der Salafismus, dessen Weltanschauung am Vorbild der Vorfahren orientiert ist. Hier kann allerdings beobachtet werden, dass mit der Vergegenwärtigung von Symbolen und Erinne                                                                                                                 61 Q 2:129: „Unser Herr! Lass unter ihnen einen Gesandten erstehen, aus ihrer Mitte, der ihnen deine Verse vorträgt, sie das Buch und die Weisheit lehrt und sie läutert! Siehe, du bist der Mächtige, der Weise.“ 62 Vgl. Jane Dammen McAuliffe, „The Prediction and Prefiguration of Muḥammad“, in: John C. Reeves (Hg.): Bible and Qurʾān, Leiden: Brill 2004, S. 107-131. 63 Q 2:124-132. 64 Vgl. Uri Rubin, „Between Arabia and the Holy Land: a Mecca-Jerusalem axis of sanctity”, in: Jerusalem Studies in Arabic and Islam 34 (2008): 345-362.

360

Georges Tamer

rungen aus der tiefen Vergangenheit eine Authentizität der Identität beansprucht wird, die vor allem in konflikthaften Situationen ausschließenden Charakter besitzen kann.

Aspect in the Slavic and the Biblical Hebrew imperative Silje Susanne Alvestad, University of Oslo, and Lutz Edzard, University of Erlangen Nürnberg and University of Oslo

1

Introduction

In this joint paper, we attempt to explore to which degree the advanced discussion on aspect in the Slavic languages can be applied in a meaningful way to the corresponding discussion in Semitic, more specifically to Biblical Hebrew injunctive (modal deontic) forms. Scholars working on the Slavic verbal system have certainly be aware of grammatical features in Semitic (cf., e.g., Fortuin 2000: 144) and scholars working on verbal systems in Semitic at times refer the the discussion in Slavic and theoretical linguistics (cf., e.g., Boneh 2013, referring to Filip 2011). Henkin (1994: 277–283), referring to earlier work by Heikki Palva (1977, 1984), compares the so-called narrative imperative found in Arabic Bedouin dialects in the Negev with the “dramatic” or narrative imperative as attested in Russian, all this in a purely descriptive way. In the first section of this paper (sections 2 and 3), we will look more closely at positive (imperative) and negative (prohibitive) commands in Biblical Hebrew. Of special interest to us is the question as to whether the opposition between perfective aspect, usually expressed by imperative and jussive forms, and imperfective aspect, usually expressed by the indicative (“imperfect”) is always encoded in this way, or whether there exist exceptions, especially in the negated imperative (prohibitive). 2

Summary of the conventional wisdom

2.1

Summary of basic morpho-syntactic and semantic facts

The Biblical Hebrew finite verb basically comes in two types of conjugation: ̄ vC/, which (i) a suffix conjugation designating mainly past with the base /CåC ̄ vC/ (third person sg. m.) – the also underlies the consecutive non-past /wә-CåC latter with final stress in the first and second person singular; and (ii) a prefix conjugation designating mainly non-past (indicative; also modal/ volitional) with the base /CCvC/, which also underlies the jussive and the

362

Silje Susanne Alvestad and Lutz Edzard

imperative moods, as well as the consecutive past /way-yiCCvC/ (third person sg. m.) – the latter with penultimate stress, if the penult is open. The jussive overlaps morpho-phonologically with the consecutive past (minus the waC-prefix); disregarding differences in stress, both the jussive and the consecutive past are only distinguishable from the (indicative) prefix-conjū (hip̄ ʿīl). Here are the basic gation in the case of weak verbs and the H-binyån paradigms of the verb “to write” (in the Semiticists’ tradition beginning with the third person) (1): (1)

The basic conjugations in Biblical Hebrew

ps. 3ms

suffix conjugation (past) kå̄ṯaḇ kå̄ṯәḇå̄(h) kå̄ṯaḇtå̄ kå̄ṯaḇt kå̄ṯaḇtī kå̄ṯәḇū

3fs 2ms 2fs 1cs 3mp 3fp 2mp 2fp 1cp

kå̄ṯәḇū kәṯaḇtεm kәṯaḇtεn kå̄ṯaḇnū

prefix conjugation (non-past) yiḵtōḇ tiḵtōḇ tiḵtōḇ tiḵtәḇī ʾεḵtōḇ yiḵtәḇū tiḵtōḇnå̄(h) tiḵtәḇū tiḵtōḇnå̄(h) niḵtōḇ

There exist phonological differences between the past and and the consecutive non-past forms on the one hand – the difference simply lies in stress (penultimate vs. ultimate) – and differences between plain indicative forms and consecutive past and jussive forms on the other hand (cf., e.g., Blau 1993: 47). In the case of the jussive and consecutive past forms, vowel change (shortening or elision takes place (2): (2)

Differences between indicative forms and consecutive past (preterite) ̄ (hip̄ ʿīl)) and jussive forms (mainly weak verbs and verbs in the H-binyån yå̄qū́m ‘he rises/will rise’ vs. ̄ yåqṓm ‘let him rise’ (jussive) and way-yǻ ̄qåm ‘and he rose’ √ q–w–m yå̄šīṯ yå̄šēṯ way-yå̄šēṯ

‘he sets up’ vs. ‘he set up’ (old preterite form: Ps 18:12) and ‘and he set up’ (2 Sam 22:12)

√ š–y–t

Aspect in the Slavic and the Biblical Hebrew imperative

363

yaʿăśε(h)

‘he does / will do’ vs.

way-yaʿaś

‘and he did’

yәḇå̄rḗḵ wa-yḇǻ ̄rεḵ

‘he blesses/will bless’ vs. ̄ ( qatl-u (Steiner 2012: 367); 3) A vocalic metathesis based on the sg. construct, which results in the movement of the vowel in the second syllable to the initial syllable, thus creating a pseudo-segholate: ***zәqan > zaqn- (Joüon and Muraoka 2000: I96 Bd).

2

This form is not an adjective; contra Steiner (2012: 369).

3

An identical phrase in plural is found in Ezek. 44:7.

4

The only absolute adjectival form attested for this root is ʾǎrōk; however, there is no alternation between *qatul and *qatl in any other context. It is, therefore, probable that ***ʾārēk is simply not attested or was lost. The construct ʾerek is attested 15 times, 14 of which as ʾerek ʾapayim, which is a fixed phrase.

428

Naʿama Pat-El

As was noted above, Steiner (2012) claims that the morphological variation is a result of syncope, in which case explanation 1 and 2 are one and the same, but I will treat them here as two separate explanations. The first proposal is problematic because it may account for the substantives, but it cannot account for the adjectival forms, where qatl is not a possible alternate pattern. The second proposal is problematic because such a syncope is ad hoc. The third proposal is problematic because it is unmotivated, and unknown elsewhere in Hebrew. I suggest a different explanation, which accounts for both substantives and adjectives and is based on regular sound rules in Hebrew. The plural formation of qatil nominals, e.g., kәbēdīm (< *ka-bi-dīm), is a natural result of the reduction of short non-pretonic vowels in open syllables, and the lengthening of short pretonic vowels in open syllables. The construct plural, however, has two short pretonic vowels in an open syllable, *ka-bi-day, which are therefore reduced to schwa, *kәbәdē. Such a sequence is regularly broken by a high vowel in what is known as “the rule of schwa”: *kabiday > *kәbiday > *kәbәdē > kibdē. This construct form is identical to the plural construct of qitl nouns; e.g., birkē ‘knees’ (abs. berek < *birk), ḥiṣṣē ‘arrows’ (abs. ḥēṣ < *ḥiṣṣ). The construct form is interpreted by speakers to be from the base qitl, which leads to a reanalysis and a production of a singular construct bi-form, such as kebed etc. Steiner’s claim that these adjectives are very archaic is not borne out by the facts. The evidence is rather scant, but at least for the adjectives ʿārēl and kābēd, the earliest forms clearly derive from *qatil, while the later forms in Ezekiel, from *qVtl. Given the limited number of examples, I am not sure we can say anything conclusive about the date of these forms in Hebrew. The sequence of rules I have outlined above is not attested in other languages, and indeed other languages did not develop adjectives belonging to the *qVtl pattern. This explanation accounts for the variation in both adjectives and substantives without resorting to special explanations which are either rare or ad hoc. Additionally, it also explains why the segholate bi-form is exclusively found in construct.

A note on segholate adjectives in Biblical Hebrew

429

References Fox, Joshua. 2003. Semitic Noun Patterns. Winona Lake: Eisenbrauns. Huehnergard, John. 2013. “Segholate: Pre-Modern Hebrew.” In: Encyclopedia of Hebrew Language and Linguistics, ed. Geoffrey Khan, vol. 3, 520-522. Leiden: Brill. Joüon, Paul and Takamitsu Muraoka. 2000. A Grammar of Biblical Hebrew. Rome: Editrice Pontificio Instituto Biblico. Lambdin, Thomas O. 1985. “Philippi’s Law Reconsidered.” In: Biblical and Related Studies Presented to Samuel Iwry, ed. by Ann Kort and Scott Morschauser, 135-145. Winona Lake: Eisenbrauns. Militarev, Alexander and Leonid Kogan. 2000. Semitic Etymological Dictionary. Vol. 1. Anatomy of Man and Animals. Münster: Ugarit Verlag. Rabin, Chaim. 1951. Ancient West Arabian. London: Taylor’s Foreign Press. Steiner, Richard. 2012. “Vowel Syncope and Syllable Repair Processes in Proto-Semitic Construct Forms: A New Reconstruction Based on the Law of Diminishing Conditioning”. In: Language and Nature: papers presented to John Huehnergard on the Occasion of his 60th Birthday, ed. by Rebecca Hasselbach and Naʿama Pat-El, 365390. Chicago: The Oriental Institute of the University of Chicago.

Eliʿezer Ben-Yehuda and Algerian Jews: Relationship and language* Ofra Tirosh-Becker, The Hebrew University of Jerusalem

1. Introduction Eliʿezer Ben-Yehuda (1858-1922) was a pioneer in the revival of spoken Hebrew. He composed the most comprehensive Hebrew dictionary of his time, entitled The Complete Dictionary of Ancient and Modern Hebrew, in which he documented Hebrew vocabulary from all strata of the Hebrew language: biblical, rabbinic, medieval, and modern.1 Ben-Yehuda’s ideology for the revival of Hebrew speech was disseminated through Hebrew language journals, some of which he edited and published himself.2 One facet of Ben-Yehuda’s life and work that had hardly been noted so far is the relationship between Eliʿezer Ben-Yehuda and Algerian Jews. In this paper we will discuss three aspects of this relationship. The first aspect is BenYehuda’s sojourn in Algeria in the winter of 1880/81,3 the second is the stature of Ben-Yehuda in the eyes of Algerian scholars as manifested in the eulogies

*

This paper is based on a lecture delivered at the 44th annual conference of the Association for Jewish Studies, Chicago, IL, USA on December 16, 2012, which marked exactly the 90th yahrzeit of Eliʿezer Ben-Yehuda, who passed away on December 16, 1922. As a researcher of the Hebrew language, and in particular as I was personally acquainted with Ben-Yehuda’s daughter, Dola Ben-Yehuda Wittmann (1902–2004), it was especially inspiring for me to lecture on Ben-Yehuda on that particular day. This study was supported by a grant from the Israel Science Foundation (578/13).

1

Ben-Yehuda 1948.

2

Ben-Yehuda was the editor of several Hebrew-language newspapers: Ha-Zvi, HaʾOr, and Hashkafa. These journals can be read online on The Historical Jewish Press site at http://web.nli.org.il/sites/JPress/Hebrew/Pages/AllJPressPage.aspx.

3

Ben-Yehuda’s sojourn in Algeria was noted in: Harshav 1990, p. 39; Y. Charvit, The History of Algerian Jewry during the French Period (1830-1962), Tel Aviv 2010 [Hebrew], p. 81; O. Tirosh-Becker, “Algeria”, in Encyclopedia of Hebrew Language and Linguistics, Ed. G. Khan, Brill Academic Publishers: Boston, 2013, vol.1, pp. 85-86.

Eliʿezer Ben-Yehuda and Algerian Jews: Relationship and language

431

published in Algeria following his death, and the third is the impact of BenYehuda’s language innovations on modern Hebrew in Algeria. 2. Ben-Yehuda’s sojourn in Algeria The relationship between Eliʿezer Ben-Yehuda and Algerian Jews began in the winter of 1880/81, which Ben-Yehuda spent in Algeria at the advice of his physicians, just a few months before his immigration to the Land of Israel. In the introduction to his monumental dictionary Ben-Yehuda reports of the great impression that the interaction with the Algerian elders had made on him. It was in Algiers, where he first heard Jews reading the Torah in Sephardic pronunciation. Furthermore, it was there that he had, for the first time, conversed in Hebrew out of actual necessity, and not for the sake of the idea of speaking in Hebrew. Ben-Yehuda conveys his excitement over the Hebrew conversations he held with the elders of Algiers’ Jewish community, and reports that at times it felt like his natural speech. To quote from Ben-Yehuda’s introduction to his dictionary:4 Meanwhile I suffered from tuberculosis, had to quit my studies in medical school, and was sent by order of my physicians to the city of Algiers. It was there that I heard for the first time Jews reading the Torah in a Sephardic pronunciation. That pronunciation left a very strong impression on me. It was in Algiers that for the first time I spoke Hebrew not for the sake of speaking Hebrew but out of actual necessity, because I could speak with their elders and ḥakhamim (= scholars), who did not know French, only in leshon haqodesh (= the holy tongue, i.e. Hebrew), which some of them mastered also as a language of speech. The days that I spent in Algiers bore double blessings. The African sun healed my body, and my Hebrew conversations with the elders of the Israelite community and its ḥakhamim improved my Hebrew speech proficiency. Moreover, there were even times when I felt that speaking Hebrew is my natural tongue. During his stay in Algiers Ben-Yehuda published an open letter to Peretz Smolenskin, the editor of the Hebrew journal Ha-Shaḥar (= The dawn),5 in which he outlined his belief that the renewal of the Jewish nation in its homeland is coupled to the revival of Hebrew as the language of speech for its masses. This

4

Ben-Yehuda 1948, Prolegomena, p. 5.

5

Ha-Shaḥar (= The dawn), vol. 10, issue 5 (1881); see E. Ben-Yehuda, Collected Works, Vol. 1, Jerusalem 1941, pp. 27-33 [Hebrew].

Ofra Tirosh-Becker

432

letter, sent from Algiers, was one of the very first articles in which Ben-Yehuda formulated and conveyed his ideology.6 Ben-Yehuda’s experience in Algiers, and his exposure to the Algerian Sephardic pronunciation of Hebrew, had no doubt impacted his subsequent efforts towards the revival of Hebrew as a spoken language, as well as his support of choosing the Sephardic pronunciation for the revived Hebrew speech.7 As Ben-Yehuda himself describes his impression of this pronunciation in explicit words: va-taʿas ʿalay havara zo rošem ʿaz meʾod (‫= ותעש עלי הברה זו רושם עז מאוד‬ and that [i.e., Sephardic] pronunciation left a very strong impression on me).8 The aforementioned quotation from Ben-Yehuda’s prolegomena also offers a rare evidence that Hebrew was in fact spoken on special occasions by the Rabbanite elite in late nineteenth century Algeria, at least when meeting a foreign Jewish traveler. This counters the common belief that prior to its revival Hebrew had only a literary existence. The present testimony augments other evidence of this kind that Jewish travelers, when visiting far off Jewish communities, resorted to Hebrew as their language of oral communication. For example, in the epilogue to his book Maḥberet He-ʿArukh, the twelfth century Shelomo ben Farḥon writes that Jews from different European countries used Hebrew for inter-communication because of the mutual unintelligibility of their local languages.9 Likewise, the German traveler Arnold von Hertz reported that in the late fifteenth century Hebrew was spoken in Jerusalem, where Jews from a mosaic of communities lived side by side.10 Finally there is evidence that in

6

This is Ben-Yehuda’s second article on the subject. The first article, entitled Degel ha-leʾumiyyut (= Banner of nationalism), was written in Paris in late 1880. Interestingly, in his famous article Sheʾela Nikhbada (= A weighty question) Ben-Yehuda does not actually discuss the revival of Hebrew but only hints to the possibility of informal Hebrew speech. See Eldar 2010, pp. 40-41, fn. 60, 65 and references therein. Also see R. Kuzar, Hebrew and Zionism: A Discourse Analytical Cultural Study, Berlin: Mouton de Gruyter 2001, pp. 41-136. 7 For his 1903 speech in support of the Sepharadic pronunciation see: ‫לקט תעודות‬ ‫ ולחידוש הדיבור העברי‬,‫ תר"ן – תש"ל‬:‫לתולדות ועד הלשון‬, The Academy for Hebrew Language: Jerusalem, 1970, pp. 160-161 [Hebrew]. Also see Eldar 2010, pp. 59-62; Harshav 1990, pp. 37-44. 8 See fn. 4 above. 9 Morag 1957, pp. 9-10.

10 Morag 1957, p. 10; J. Klausner, “‫ =( ”הדבור העברי בא"י במאה הט"ו‬Hebrew speech in Eretz Yisrael in the 15th century), in A. Tzifroni, A. Z. Rabinovich, and D. Shimʿon-

Eliʿezer Ben-Yehuda and Algerian Jews: Relationship and language

433

Yemen and Morocco rabbinic scholars used Hebrew for their scholarly discourse.11 3. The stature of Ben-Yehuda in the eyes of Jewish Algerian scholars Ben-Yehuda’s respect for the Algerian Jewish community was reciprocated with admiration. Ben-Yehuda was revered by the Algerian circles of maskilim, scholars of the Jewish Enlightment movement.12 On the stature of Ben-Yehuda among Jewish maskilim in Algeria we learn from eulogies for Ben-Yehuda in the weekly journal ǝl-Ḥikma, which was the only Judeo-Arabic journal published in Algeria in the first half of the twentieth century.13 El-Ḥikma was printed in Constantine (the third largest city in Algeria) under the editorship of Rabbi Avraham Zerbib.14 According to publications in this journal, the passing of Eliʿezer BenYehuda on December 16, 1922 was considered – at least in this Algerian Jewish community – a major event of national proportion. An obituary signed by the journal’s director and employees was published at the top of the January 5, 1923 issue. The obituary reports on the demise of “the reviver of the Hebrew lan-

ovitch (eds.), ‫ =( ספר השנה של ארץ ישראל‬The Yearbook of Eretz Israel; Tel Aviv, 1923), pp. 114-117. 11 Morag 1957, p. 10. 12 Ben-Yehuda was also admired in other places in North Africa. In Libya an Eliʿezer Ben-Yehuda Society was established in 1931, with the goal of advancing Hebrew speech, see S. Sela, “Education”, in Libya: Jewish Communities in the East in the Nineteenth and Twentieth Centuries, ed. H. Saadoun, Jerusalem: Ben-Zvi Institute 2007, pp. 105-108 [Hebrew]. Also see Sivan 1973, pp. 113-114. 13 After 1896 all other Jewish journals in Algeria were published in French, while Jewish journals published in nineteenth century Algeria prior to 1896 were written in Judeo-Arabic or appeared as bi-lingual Judeo-Arabic and French publications. See: Tirosh-Becker 2011, pp. 130-132. 14 Rabbi Avraham Zerbib (1870-1942), a native of Constantine, held an important position in his community as the chief ritual slaughterer. Later, in the 1930’s, he became the Chief Rabbi of a neighboring town Setif where he served for many years, returning to Constantine only in old age. Zerbib was both a religious leader and a supporter of the Jewish Enlightment movement (Haskala). In addition to editing the journal El-Ḥikma, Zerbib composed several treatises that were published only after his death. See Marciano 2002, p. 308; Y. Charvit, Elite rabbinique d'Algérie et modernisation, 1750-1914, Jerusalem, 1995, p. 101.

434

Ofra Tirosh-Becker

guage in the Land of Israel”.15 This obituary was followed by a series of no less than five eulogies for Ben-Yehuda that appeared in subsequent issues of the journal. The first of these eulogies appeared on January 12, 1923, in the first issue of ǝl-Ḥikma published following the issue with the obituary. This eulogy, composed by the journal’s editor Rabbi Avraham Zerbib in Judeo-Arabic, sheds light on Ben-Yehuda’s standing among Algerian maskilim. In his eulogy Zerbib explains why he entitled Ben-Yehuda ‘the reviver of the Hebrew language’ and not ‘the father of the Hebrew language’, as he was designated in the journal Doʾar Ha-Yom (= Daily Post, Jerusalem, 1919–1936) published in the Land of Israel by Itamar Ben-Avi, Ben-Yehuda’s own son. According to Zerbib the appellation ‘father of the Hebrew language’ was given to Ben-Yehuda in the Land of Israel, because he ‘gave birth’ of new Hebrew words.16 Namely, his lexical innovations position him as a father of the Hebrew language. However, for Algerian Jews Ben-Yehuda was the ‘reviver of the Hebrew Language’, since Jews in the Diaspora spoke local languages at the expense of Hebrew, until BenYehuda demanded of them to speak Hebrew for any and every purpose.17 In that issue, and in the two subsequent ones, the editor published JudeoArabic translations of the Hebrew eulogies that were read during Ben-Yehuda’s funeral. These eulogies were composed by Menaḥem Ussishkin (1863–1941; a Zionist leader and head of the Jewish National Fund), David Yellin (1864–1941; a Zionist leader, among the founders of the Hebrew Language Committee and the Jewish Teachers Federation), Yosef Meyuḥas (1868–1942; founder of the National Library of Israel and among the founders of the Hebrew Language Committee) and Dr. Aharon Meʾir Mazia (1858–1930; a physician and a linguist, among the founders of the Hebrew Language Committee). All four were BenYehuda’s colleagues in the herculean task of reviving the Hebrew language. According to the journal’s editor, Rabbi Avraham Zerbib, he translated these

15 The original wording in Judeo-Arabic: ‫מוחיי אללסאן אלעברי פי אלוטן ארץ ישראל‬. This title was not unanimously accepted by Ben-Yehuda's contemporary opponents, see Reshef 2014, pp. 613-614. 16 The original wording in Judeo-Arabic: ‫לי יתסמא אב הלשון ראה יולד לכלאם ללסאן לעברי‬. 17

‫ והאד אלכייס ואקף‬,‫ ולסאן לעברי תנסא מנהום‬,‫מן סבת ליהוד פלגלות צארו יתכלמו בלסאן לווטן פאיין ראהום‬ ‫ =( יללזם )!( ליהוד יתכלמו בלסאן לעברי פי כל מסאלה‬Because the Jews in the Diaspora spoke the language of their country they had forgotten the Hebrew language, and this smart man demanded of the Jews to speak in Hebrew for all purposes).

Eliʿezer Ben-Yehuda and Algerian Jews: Relationship and language

435

eulogies from the December 19, 1922 issue of the Jerusalem-based journal Doʾar Ha-Yom, in which they were published a mere three days after Ben-Yehuda’s departure. In fact, Ben-Yehuda was considered by Algerian maskilim as one of the cornerstones of Zionism, on the same pedestal as Herzl. This is well described in a Judeo-Arabic article written by Avraham Zerbib on March 23, 1923, three months after Ben-Yehuda’s demise. In that article the author lists ‘four redemptions’ (ʾarbaʿ gәʾullot) based on the four verbs in Exodus 6:6–7,18 which are entitled the ‘four languages of redemption’ (ʾarbaʿ lәšonot gәʾulla). These verbs are vә-hoṣeti (= I will lead you out), vә-hiṣalti (=I will save you), vә-gaʾalti (= I will redeem you), and vә-laqaḥti (= I will take you). Zerbib also lists four redeemers (ʾarbaʿa goʾalim), whose memory should be cherished forever, and four redeemed (ʾarbaʿa gәʾulim), which are the foundations of the Zionist movement. These four are Theodor Herzl19 – ‘the redeemer of the nation’, Joseph Chasanowich – ‘the redeemer of the [Hebrew] literature’,20 Eliʿezer Ben-Yehuda – ‘the redeemer of the [Hebrew] language’, and Max Nordau – ‘the redeemer of the land’.21 Thus writes Zerbib about Ben-Yehuda, the redeemer of the language: Ben-Yehuda claimed that it was not sufficient for Jews to live in the land of Israel and speak foreign languages. It is essential that Jews in their homeland will speak only Hebrew at all times and for all purposes. He took upon himself that he and his family will speak only Hebrew. He taught his wife Hebrew and insisted that she speaks with her children only in Hebrew. It is told that once Ben-Yehuda returned home and saw his wife rocking their son to sleep singing in Russian. He was angry at her, and instructed her to put the baby to sleep using the Hebrew words šәxav bәni, šәxav bәni, meaning ‘lay down to sleep my son’. On another occasion Ben-Yehuda shared with his friend that he and his wife would like their baby to speak only Hebrew. The friend responded saying: “You are crazy (ʾanta mahbul) teaching your son a dead language. What good will come of this? The child will be like a madman himself.” However, Ben-Yehuda adhered to his mission; his son spoke only Hebrew, and did not become a madman… Today he is 18 Exodus 6:6–7 is part of the Torah section known as parashat va-ʾEra, 19 On Theodor Herzl (1860–1904) see I. Friedman, "Herzl, Theodor", EJ2, Vol. 9, pp. 54-66. 20 On Joseph Chasanowich (1844-1919) see A. Yaari, "Chasanowich, Joseph", EJ2, Vol. 4, pp. 580-581. 21 On Max Nordau (1849-1923) see M. Ben-Horin, "Nordau, Max", EJ2, Vol. 15, pp. 297-299.

436

Ofra Tirosh-Becker

the chief editor of the large Jerusalemite journal Doʾar Ha-Yom that is written in this precious language, which is no longer a dead tongue. BenYehuda revived the Hebrew language, and even non-Jews among the nations started speaking it, as we described in length in previous issues of the journal ǝl-Ḥikma. It is said of Ben-Yehuda vә-gaʾalti (= and I will redeem) because he is the one who redeemed the Hebrew language from oblivion (lit. from its death). 4. Hebrew language innovation in Algeria Circles of maskilim, namely scholars of the Jewish Enlightenment movement, were active in Algeria in the second half of the nineteenth century. Their activity included organization of Hebrew cultural events, establishment of a society for advancing Hebrew studies, and founding of public libraries thus making Modern Hebrew literature and journals accessible to the public.22 Among the leading maskilim in Algeria were Shalom Bekache23 and Isaac Morali of Algiers,24 and Ḥaim Beliaḥ of Tlemcen.25 The Algerian maskilim were also the driving force behind the development of Hebrew press in Algeria towards the

22 Chetrit 1993, pp. 90-123; Chetrit 1990, p. 26; J. Chetrit, “New Consciousness of Anomaly and Language: The Beginnings of a Movement of Hebrew Enlightment in Morocco at the end of the Nineteenth Century”, Miqqedem Umiyyam 2, Studies in the Jewish Society in Islamic Countries and Sephardic Diaspora (1986), pp. 129-168 [Hebrew]; J. Chetrit, “La Haskala hébraïque au Maroc à la fin du XIX siècle et sa Contribution à l’apparition do mouvement sioniste”, in I. Ben-Ami (ed.), Recherches sur la culture des juifs d’Afrique du Nord, Jerusalem 1991, pp. 313-331 [Hebrew]. 23 Y. Tobi, "Bekache, Shalom”, EJIW, vol. 1, pp. 368-369. Bekache published in his Judeo-Arabic newspaper Beth Yisrael (its French title: Le peuple d’Israel) an essay praising the Hebrew language and its role in Jewish culture (issue 1 [June 25, 1891], cols. 4-9; issue 2 [July 2, 1891], cols. 21-25). See Tirosh-Becker 2011, p. 129; Chetrit 1993, p. 101; Chetrit 1990, pp. 18, 27, 40-42. 24 Y. Charvit. "Morali, Isaac”, EJIW, vol. 3, pp. 458-459; E. Hazan, “The Literary Activity of R. Isaac Morali and his Plan to Collect the Hebrew Poetry of Algeria”, Peʿamim 91 (2002), pp. 65-78 [Hebrew]; Chetrit 1990, pp. 42-43. 25 Marciano 2002, p. 134; Y. Charvit, “ ‫( – רבה של תלמסאן וחכמי‬1832-1919 ) ‫הרב חיים בלייח‬ ‫ הרחבת תקנת הפקעת הקידושין‬:‫ =( ”ארץ ישראל‬Rabbi Ḥaim Beliaḥ (1832-1919) the Rabbi of Tlemcen and Eretz Yisrael), Mahut: Journal of Jewish Literature and Art 25 (2002), pp. 163-174 [Hebrew]; E. Hazan, “ – ‫איגרת מרבי חיים בלייח מתלמסאן לרבי אשר ג'אמי בתוניס‬ (1881) ‫ =( ”התרמ"א‬A letter from Rabbi Ḥaim Beliaḥ of Tlemcen to Rabbi Asher Jami in Tunis 1881), Studies in the Culture of North African Jewry: Edited and Interpreted Texts, M. Bar-Asher and S. Fraade (eds.), Jerusalem 2011, pp. 107-113 [Hebrew].

Eliʿezer Ben-Yehuda and Algerian Jews: Relationship and language

437

end of the 19th century.26 The translation of Avraham Mapu’s pivotal book ʾAhavat Zion (= The Love of Zion) into Algerian Judeo-Arabic by Rabbi Yosef Renassia (1879-1962) of Constantine27 testifies to the importance of Hebrew Enlightenment literature to Algerian Jews. Interestingly, Mapu’s book was also translated into other Judeo-Arabic dialects28 as well as other Jewish languages, such as Yiddish,29 Ladino,30 and Judeo-Persian.31 These maskilim, for whom the Hebrew language was an integral part of their national identity, submitted Hebrew articles to European Jewish Haskala newspapers that were read in urban centers throughout North Africa, some even

26 R. Attal and M. Harroch, “Hebrew and Judaeo-Arabic Printing in Algiers”, Kiryat Sefer 61:3 (1986-7), pp. 561-572 [Hebrew]; R. Attal, “Hebrew and Judaeo-Arabic Printing in Oran (Algeria)”, Kiryat Sefer, Collected Essays, supplement to vol. 68 (1998), pp. 85-92 [Hebrew]. See also J. Fraenkel, L’Imprimerie Hébraïque à Djerba (étude bibliographique), Thèse de doctorat de troisième cycle, Universite de Paris III, Paris, 1982; R. Attal, “The Books of Shalom Békache, Publisher in Algiers”, Alei Sefer 2 (1976), pp. 219-228 [Hebrew]; Cf. J. Tedghi, Le livre et l'imprimerie hébraïques a Fes, Jerusalem: Ben-Zvi Institute 1994 [Hebrew]. 27 O. Tirosh-Becker, Phonology and Topics in the Morphology of a Judeo-Arabic Translation of Psalms from Constantine (Algeria), Master’s thesis, The Hebrew University of Jerusalem, 1988, pp. 4-6 [Hebrew]; Y. Charvit, “Renassia, Joseph,” EJIW, vol. 4, pp. 157–159; Y. Charvit, “Rabbi Renassia: A Portrait of a Spiritual Leader in French Algeria (1879-1962)”, in M. Orfali and E. Hazan (eds.), Progress and Tradition: Creativity, Leadership and Acculturation Processes among the Jews of North Africa, Jerusalem 2005, pp. 89-96 [Hebrew]. 28 R. Attal, “Les traductions en judéo-arabe tunisien des oeuvres d’Abraham Mapu“, Revue des Etudes Juives 134 (1975), pp. 137-144. For example, Zemaḥ ben Nathan Halevi, ‫ =( אהבת ציון או חכאית אמנון ותמר‬The Love of Zion or the Story of Amnon and Tamar), Tunis, [before 1890]. 29 For example, Menahem Berish Appleboim, ‫ =( אמנון און תמר‬Amnon and Tamar), Warsaw 1923. 30 For example, David Fresko, ‫ רומאנסו דיל איסקריטור אברהם מאפו‬:‫ =( אמור די ציון‬The Lovers of Zion: A Romance by Avraham Mapu), Saloniki 1894. On the author see David M. Bunis, “The Autobiographical Writings of Constantinople Judezmo Journalist David Fresco as a Clue to His Language Attitudes,” C. Herzog & R. Wittmann (eds.), Self-Narratives of the Ottoman Realm: Individual and Empire in the Near East, vol. 1, London: Ashgate, forthcoming. 31 For example, Shimeʿon Ḥakham, ‫ =( ספר אהבת ציון‬The Book of the Love of Zion), Jerusalem 1913.

438

Ofra Tirosh-Becker

serving as their local correspondents.32 As early as 1864 Ha-Maggid (= The Preacher; the first Hebrew weekly journal published in central Europe, 1856– 1903) published an article by Eliyahu Djerbi of Blida, a town near Algiers.33 It was followed by a series of articles from a variety of Algerian communities. Eulogies for North African figures of stature were also published in these journals, indicating that they had a relatively broad readership in North Africa.34 Reports from North Africa also appeared in Hebrew journals published in Jerusalem in the ‘80s and ‘90s of the nineteenth century, such as Ben-Yehuda’s newspapers Ha-ʾOr (= The Light, Jerusalem, 1890–1893) and Ha-Zvi (= The Gazelle, Jerusalem, 1884–1902, 1908–1915).35 Despite the broad Jewish journalistic activity in Algeria, no Hebrew newspaper was published in Algeria itself. Early Algerian Jewish newspapers, the first of which was the bi-lingual French and Judeo-Arabic L’Israélite Algérien/әddziri (= The Algerian) in 1870, were published either in Judeo-Arabic or as bilingual French and Judeo-Arabic journals, while later journals were mainly published in French alone.36 It is through Hebrew journals that Ben-Yehuda’s lexical innovations reached the Algerian maskilim, in particular via Itamar Ben-Avi’s journal Doʾar Ha-Yom that frequently published these innovations.

32 Chetrit 1990, pp. 13-14; Chetrit 1993, pp. 111-112. For example, Hebrew Haskala journals were read by Jewish maskilim in Morocco, some of which sent articles for publication in these journals, e.g. Isaac ben Yaʿish Halevi from Mogador, who sent articles to Ha-Zǝfira and Ha-Maggid. See A. Maman, “Language”, in Morocco: Jewish Communities in the East in the Nineteenth and Twentieth Centuries, ed. H. Saadoun, Jerusalem: Ben-Zvi Institute 2003, pp. 150-151 [Hebrew]. 33 Chetrit 1990, pp. 13, 55 fn. 8. 34 For example, eulogies for Rabbi Refaʾel David Morali (November 25, 1892, by Shalom Bekache) and for Yossef Ṣerur (February 1, 1894, by Isaac Morali) were published in Ha-Maggid. Cf. Chetrit 1990, p. 56 fn. 16. 35 See footnote 2 above. 36 R. Attal, “The first Jewish newspaper in the Maghreb - L’Israélite Algérien, 1870”, Peʿamim 17 (1984), pp. 88-95 [Hebrew]; R. Attal, Périodiques juifs d’Afrique du Nord, Jerusalem: Ben-Zvi Institute, 1980 [Hebrew]; R. Attal, La presse périodique juive d’Afrique du Nord, Tel-Aviv 1996 [Hebrew]; Tirosh-Becker 2011, pp. 130-132. On the use of French by Algerian Jews see C. Aslanov, “The French Spoken by Algerian Jewry,” in Algeria: Jewish Communities in the East in the Nineteenth and Twentieth Centuries, ed. H. Saadoun, Jerusalem: Ben-Zvi Institute 2011, pp. 133-138 [Hebrew]. See also M. Cohen, Le parler arabe des juifs d’Alger, Paris 1912, pp. 1-15.

Eliʿezer Ben-Yehuda and Algerian Jews: Relationship and language

439

The modernity that was introduced by French colonization required Algerian Hebrew and Judeo-Arabic scholars to address an array of new concepts, technologies and discoveries in multiple and rapidly developing fields of knowledge. Thus it was necessary to expand the lexicon of both Hebrew and Judeo-Arabic accordingly. An effort towards that goal is reflected in a trilingual French-Hebrew-Arabic dictionary that was published by Rabbi Yosef Renassia of Constantine around 1930.37 This comprehensive trilingual dictionary holds more than 11,000 entries, reflecting a broad range of topics from all walks of life. In this dictionary Rabbi Renassia relied on his mastery in the various strata of Hebrew. For example the entry yeter ʿoz (‫ )י ֵתֶ ר עָז‬for ‘reinforcement’ was drawn from biblical Hebrew,38 while ʾeven šoʾevet (‫ ) ֶאבֶן ש ֹ ֶאבֶת‬literally ‘a drawing stone’ for ‘a magnet’ is based on its use in the Talmud.39 The phrase ḥoxmat hanefeš ve-xoḥoteha (‫ ; ָח ְכ ַמת ַהנֶּפֶש ְוכֺחוֹתֶ י ָה‬lit. the wisdom of the soul and its fortitude) that denotes ‘psychology’,40 is coined in a template that was common in medieval Hebrew, similar to ḥoxmat ha-lašon (‫ ) ָח ְכ ַמת ַה ָלשוֹן‬that reflects the Arabic ʿilm al-luġa (‫)ﻋﻠﻢ ﺍاﻟﻠﻐﺔ‬.41 Both ḥoxmat ha-lašon and ʿilm al-luġa for ‘lexicography, philo-

37 Renassia 1930. See Tirosh-Becker 2011, pp. 129-130. Each entry in this trilingual dictionary includes five columns: 1. French, 2. Hebrew in Hebrew script, 3. Hebrew transcribed into Latin characters, 4. Arabic in Arabic script, 5. Arabic transcribed into Latin characters. I intend to publish a comprehensive study of this dictionary elsewhere. 38 Renassia 1930a, p. 381: renfort – ‫( י ֵתֶ ר עָז‬iéthère âze) – ‫( ﺯزﻳﯾﺎﺩدﺓة ﺍاﻟﻘﻮﺓة‬ziadeth elk’oua). Cf. Gen 49:3: ‫שׂ ֵאת ְוי ֶתֶ ר עָז‬ ְ ‫י ֶתֶ ר‬. Citations from this dictionary in the present paper are given exactly as they appear in the original book (transcription, punctuation, etc.). Sometimes the Hebrew words in the dictionary entries were not fully punctuated, and at times they do not conform with standard punctuation rules. 39 Renassia 1930a, p. 12: aimant – ‫( ֶאבֶן ש ֹ ֶאבֶת‬ébène choébèthe) – ‫( ﻣﻨﻐﺎﻁطﻴﯿﺲ‬mèneghatisse). Cf. b. Soṭa 47a, b. Sanhedrin 107b, and b. ʿAvoda Zara 44a, and see Rashi’s explanation of this expression in the Talmud, e.g. in his commentary on b. Soṭa 47a: ‫אבן‬ ‫שואבת – שמגבהת את המתכת מן הארץ ומעמידתו באויר‬. The term ʾeven šoʾevet is also used by Bible exegetes such as Rabbi David Kimḥi. Cf. M. A. Ginsburg, a Jewish Enlightment scholar from Lithuania (1785-1846), added the word maḥaṭ (= niddle) to this term, forming the expression maḥaṭ ʾeven ha-šoʾevet (‫ )מחט אבן השואבת‬for ‘compass’; see Kaddari 2002, p. 38. 40 Renassia 1930a, p. 362: psychologie – ‫( חכמת הנפש וכחתיה‬h’okh’math hannéfeche ouékoh’othéha) – ‫( ﻋﻠﻢ ﺍاﻟﻨﻔﺲ‬ôlme elnefs). 41 Ben-Yehuda 1948, vol. 3, pp. 1548-1549 (entry ‫ ;)חכמה‬vol. 5, p. 2740 (entry ‫)לשון‬.

440

Ofra Tirosh-Becker

logy’ appear in this Algerian dictionary.42 The three-word expression ʾotzar kli ḥemda (‫ ;אוֹצָר ְכּלִי ֶחמְדָ ה‬lit. a collection of precious objects) for a ‘museum’ was drawn from the literature of the Haskala.43 It is also evident that the Hebrew of the Revival Period – championed by Ben-Yehuda and his colleagues – had an impact on the author of this comprehensive Algerian dictionary, as quite a few words in this dictionary can be traced back to Ben-Yehuda and his circle. It is well known that Ben-Yehuda strived to limit the influence of European languages on the revived Hebrew.44 While the earlier European maskilim often coined new terms using two- or three- word phrases that mirror the original German words, Ben-Yehuda and his circle preferred to replace those terms with

42 In his dictionary Renassia distinguishes between ‘lexicography’ and ‘philology’ (ḥoxmat ha-lašon; ‫ ) ָח ְכ ַמת ַהלָשוֹן‬and grammar (ḥoxmat ha-diqduq; ‫) ָח ְכ ַמת הַדִּ ְקדּוק‬, both coined in the same pattern. Interestingly, Renassia uses the Hebrew term ḥoxmat ha-lašon and the Arabic term ʿilm al-luġa also as the counterparts of ‘literature’. Renassia 1930a, p. 231: grammaire – ‫( ָח ְכ ַמּת הַדִּ ְקדּוּק‬sic!) (h’okh’math haddik’dok’) – ‫ﻋﻠﻢ‬ ‫( ﺍاﻟﻨﺤﻮ‬ôlme èneh’ou); p. 280: lexicographe (sic!) – ‫( ָח ְכ ַמת ַהלָּשון‬h’okh’math hallachone) – ‫( ﻋﻠﻢ ﺍاﻟﻐﺔ‬sic!) (ôlme elgha); p. 337: philologie – ‫( ָח ְכ ַמת ַהלָּשון‬h’okh’math hallachone) – ‫ﻋﻠﻢ‬ ‫( ﺍاﻟﻐﺔ‬sic!) (ôlme elgha); p. 282: littérature – ‫( ָח ְכ ַמת ַהלָּשוֹן‬h’okh’math hallachone) – ‫ﻋﻠﻢ ﺍاﻟﻐﺔ‬ (sic!) (ôlme elgh’a). 43 Renassia 1930a, p. 306: musée – ‫( אוצר כלי חמדה‬ôtsar kéli h’emda) – ‫ﺧﺰﺍاﻧﺔ ﺍاﻟﻔﻨﻮﻥن‬ (kh’zaneth el fnoune). This term is based on the biblical phrase ʾotzar kol kli ḥemda in Hosea 13:15 (‫שׁסֶה אוֹצַר כָּל ְכּלִי ֶח ְמדָּ ה‬ ְ ִ ‫ ;הוּא י‬JPS Tanakh translation: That [wind] shall plunder treasures, every lovely object). According to Ben-Yehuda’s dictionary the use of ʾotzar (‫ )אוצר‬to denote ‘a room to store objects’ is from Medieval Hebrew, see Ben-Yehuda 1948, vol. 1, p. 112 (entry ‫)אוצר‬. ʾOtzar as ‘a warehouse’ is found in a 1897 issue of Ha-Zvi, a late nineteenth century Hebrew journal from Jerusalem, see Ornan 1996, p. 7. Renassia’s use of ʾotzar in the entry musée is mirrored by his choice for the Arabic counterpart xizāna (‫)ﺧﺰﺍاﻧﺔ‬, which means ‘a warehouse’ among other things (Hava 1970, p. 166), hence xizānat ǝl-funūn means ‘the place where art objects are stored’. This pattern is also used in Algerian Arabic in the term xizānat ǝl-kutub (‫ )ﺧﺰﺍاﻧﺔ ﺍاﻟﻜﺘﺐ‬which denotes ‘a library’, namely ‘the place where books are stored’, see Beaussier 1958, p. 280; Ben Sedira 1995, p. 84. Indeed this is the Arabic term that Renassia uses for bibliothèque, see Ranassia 1930a, p. 39: bibliothèque – ‫אַרגַּז‬ ְ ‫( – ַה ְסּפ ִָרים‬argaze hassépharime) – ‫( ﺧﺰﺍاﻧﺔ ﺍاﻟﻜﺘﺐ‬kh’zaneth el kthabe). As bibliothèque means both ‘a library’ and ‘a bookcase’, Renassia in his choice for the Hebrew counterpart – ʾargaz ha-sǝfarim – refers to the latter meaning. 44 See, for example, Ben-Yehuda 1948, Prolegomena, pp. 13-14; Eldar 2010, p. 91.

Eliʿezer Ben-Yehuda and Algerian Jews: Relationship and language

441

a single Hebrew counterpart.45 Instead of mixtav ʿitti (‫ ; ִמכְתָּ ב עִתִּ י‬periodical writing) for ‘a newspaper’, which mirrors the German Zeitschrift, Ben-Yehuda coined the single word ʿitton (‫;) ִעתּוֹן‬46 instead of sefer millim (‫ ; ֵס ֶפר ִמ ִלּים‬word book) for ‘a dictionary’, which reflects the German Wörterbuch, he introduced millon (‫;) ִמלּוֹן‬47 and to denote a ‘clock’ he adopted his friend Yeḥiel Michel Pines’48 suggestion šaʿon (‫ )שָעוֹן‬instead of the two-word phrase more šaʿot (‫ ;מוֹ ֵרה שָעוֹת‬lit. hour indicator) used by the maskilim.49 Indeed, some, albeit not all, of Ben-Yehuda’s innovations found their way to Renassia’s dictionary. As the equivalent of the French word dictionnaire Renassia brings Ben-Yehuda’s word millon (‫) ִמלּוֹן‬, to which he adds in parenthesis a two-word alternative sefer hal-lašon (‫) ֵס ֶפר ַהלָּשוֹן‬, literally ‘the book of the language’, which translates his Arabic equivalent kitāb al-luġa (‫)ﻛﺘﺎﺏب ﺍاﻟﻠﻐﺔ‬.50 The fact that he felt the necessity to add an alternative to the word millon may indicate that this word was still considered rather new around 1930, when this dictionary was compiled. The Hebrew word millon occurs in this Algerian dictionary again in the term millon qaṭan (‫ = ; ִמלּוֹן ָק ָטן‬a small dictionary) as the Hebrew equivalent for the French word lexique (a lexicon). In this case the author uses the common Arabic term qāmūs (‫ )ﻗﺎﻣﻮﺱس‬as its Arabic counterpart.51 Another example is the word šaʿon that appears in this Algerian dictionary as well, although Renassia distinguishes between ‘a watch’ and ‘a clock’ as is common in French. He used the word šaʿon (‫ )שָעוֹן‬to denote ‘a watch’ (F. montre) and kept

45 Sivan 1966, pp. 189-199; D. Yellin, “ ‫”מרחיבים – מחריבים‬, – ‫ תר"ן‬:‫לקט תעודות לתולדות ועד הלשון‬ ‫ ולחידוש הדיבור העברי‬,‫תש"ל‬, The Academy for Hebrew Language: Jerusalem, 1970, p. 153 [Hebrew]. Felman 1989-1990, pp. 215-221. 46 On the word ʿitton see Sivan 1966, p. 194. 47 Sivan 1973, pp. 83-85. 48 Yeḥiel Michel Pines was a rabbi and Zionist activist (1843–1913), see G. YardeniAgmon, "Pines, Yehiel Michael," EJ2. Vol. 16, pp. 167-168. 49 Sivan 1988, pp. 14-16. An argument in favor of the new word ‫ שעון‬over the older ‫ מורה שעות‬was given by Y. Klausner, who claimed that it was necessary to distinguish between ‫ מורה שעות‬that denotes a ‘private tutor’ and ‫ מורה שעות‬that means ‘a clock’. See R. Sivan, ibid, p. 16; Sivan 1966, pp. 189-190. 50 Renassia 1930a, p. 145: dictionnaire – (‫( ִמלּון ) ֵספֶר ַהלָשוֹן‬mil-lone, séfère hallachone) – ‫( ﻛﺘﺎﺏب ﺍاﻟﻐﺔ‬sic!) (kthabe elgha). Cf. Beaussier who mentions lexicographie as one of the meanings of ‫ﻟﻐﺔ‬, see Beaussier 1958, p. 904. 51 Renassia 1930a, p. 280: lexique – ‫( ִמלּון ָקטָן‬mil-lone k’atane) – ‫( ﻗﺎﻣﻮﺱس‬k’amouss).

442

Ofra Tirosh-Becker

the older term more šaʿot (‫מוֹרה שָעוֹת‬ ֵ ) to designate horloge, namely ‘a clock’.52 For ‘train’ Renassia offers Itamar Ben-Avi’s word rakevet (‫ ) ַר ֶכּ ֶבת‬alongside the twoword alternative mesillat ha-barzel (‫) ְמ ִסלּת ַה ַבּ ְרזֶל‬, which means ‘rail’.  53 Nonetheless, there are also one-word terms coined in a similar way that had evidently not reached Algeria by 1930 or simply not adopted, as we find in Renassia’s dictionary only their two-word equivalents. For example, Renassia still uses the maskilim term mixtav ʿitti (‫ ) ִמכְתָּ ב עִתִּ י‬for ‘journal’ and not BenYehuda’s innovation ʿitton (‫) ִעתּוֹן‬.54 This two-word term is reflected in Renassia’s choice for the Arabic counterpart waraqa waqtiyya (‫)ﻭوﺭرﻗﺔ ﻭوﻗﺘﻴﯿﺔ‬, which also resonates with the Arabic term waraqa xabariyya (‫ = ;ﻭوﺭرﻗﺔ ﺧﺒﺮﻳﯾﺔ‬journal).55 Interestingly, he did not use the common Arabic terms ǧarida (‫ )ﺟﺮﻳﯾﺪﺓة‬or ṣaḥifa (‫)ﺻﺤﻴﯿﻔﺔ‬, which are documented in Algerian Muslim Arabic.56 Likewise, Renassia uses batte ʿenayim (‫ ;בָּתֵּ י ֵעינַי ִם‬lit. ‘apparatus’ for eyes) for ‘glasses’ and not mišqafayim (‫שׁ ָק ַפי ִם‬ ְ ‫) ִמ‬,57 sar ṣava (‫ ;שַׂר ָצבָא‬lit. army commander) for ‘general’ instead of maṣbi 52 Renassia 1930a, pp. 243: horloge – ‫שעות‬ ָ ‫מורה‬ ֶ (sic!) (moré chaôth) – ‫( ﺳﺎﻋﺔ ﻛﺒﻴﯿﺮﺓة‬saâ kbira); p. 303: montre – ‫( שעון‬chaône) – ‫( ﺳﺎﻋﺔ‬saâ). 53 Renassia 1930a, p. 444: train – (‫( מסלת הברזל )רכבת‬msil-lath habbarzèle-rakkébeth) – ‫( ﺷﻤﺎﻥن ﺩدﻓﻴﯿﺮ‬trik’ el h’dide, chemane dfire). Note that the term which appears in Arabic characters ‫ ﺷﻤﺎﻥن ﺩدﻓﻴﯿﺮ‬is in fact a transliteration of the French term Chemin de fer (= railway, train). Cf. bābor (‫ )ﺑﺎﺑﻮﺭر‬in Algerian Arabic, see Beaussier 1958, p. 27. The term mesillat barzel (‫ )מסלת ברזל‬appears again as part of the Hebrew equivalent for French tramway (= tram) – mesillat barzel baš-ševakim (‫)מסלת ברזל בשוקים‬, lit. a tram in the markets, see Renassia 1930a, p. 445 (interestingly, in the 3rd column of this entry Renassia wrote âghala guédoula, lit. ‘a big carriage’, instead of writing the transcription of the Hebrew term as usual). More on the word rakevet see R. Sivan, "‫ קטר ורכבת‬.‫ א‬:‫ =( "מחיי המילים‬From the life of words: a. qaṭar and rakevet), Leshonenu Laʿam 17 (1966), issue 5-6 [167-168], p. 151-153; Sivan 1973, pp. 92-93. R. Sivan reports that Aḥad Haʿam in his 1905 letter to Droyanov used the word rakevet quite unwillingly, saying that “reluctantly we have to use this Jerusalemite language” (‫)בעל כורחנו אנו צריכים ללשון הירושלמית‬, see Sivan 1966, p. 190. On the reluctance of the European maskilim to adopt the Hebrew innovations from the Land of Israel see Eldar 2010, pp. 96-100; Reshef 2014, pp. 618-623. 54 Renassia 1930a, p. 271: journal – ‫( מכְתּב עִתִּ י‬mikh’thab îtthi) – ‫( ﻭوﺭرﻗﺔ ﻭوﻗﺘﻴﯿﺔ‬ourka ouektia). Aso see Sivan 1966, p. 194. 55 Beaussier 1958, p. 1052. 56 Beaussier 1958, pp. 138 (‫)ﺟﺮﻳﯾﺪﺓة‬, 562 (‫)ﺻﺤﻴﯿﻔﺔ‬. 57 Renassia 1930a, p. 286: lunettes – ‫( בָּתֵּ י ֵענָי ִם‬batthi ênaïme) – ‫( ﻧﻮﺍاﻅظﺮ‬nouadère) [the diacritic mark of the ‫ ﻅظ‬was omitted in the text]. See Sivan 1988, pp. 11-14. The term batte ʿenayim is found already in a rabbinic responsa from the sixteenth century, see

Eliʿezer Ben-Yehuda and Algerian Jews: Relationship and language

443

(‫) ַמ ְצבִּיא‬,58 and the somewhat awkward term ‘sailor’s guide’, manhig ha-malaḥim (‫) ַמנְהִיג ַה ַמּ ָלּחִים‬, as the Hebrew equivalent of ‘compass’, rather than the newer term maṣpen (‫) ַמ ְצ ֵפּן‬, which was coined by David Yellin.59 In addition to the above tendency towards one-word terms, Ben-Yehuda’s influence can be also seen elsewhere in this Algerian dictionary. The verb ‘to fly’ (F. voler) is translated in Renassia’s dictionary by the Hebrew verb ʿaf (‫) ָעף‬. He uses the term mәʿofefa (‫ )מְעוֹ ֵפפָה‬as the Hebrew counterpart for both ‘airplane’ (F. aeroplane) and ‘airship’ (F. ballon), the latter referring to a ‘zeppelin’, which was still a dominant mean of transportation in the 1930s.60 This noun mәʿofefa resonates with the noun mәʿofef (‫ ) ְמעוֹ ֵפף‬for ‘flight pilot’, which Ben-Yehuda coined based on the Medieval Hebrew use of this word to denote ‘those whose nature is to fly’.61 For ‘airship’ Ben-Yehuda introduced the word ʾaviron (‫) ֲאוִירוֹן‬.62 N. Shapira, “‫ =( ”הלשון הטכנית בספרות הרבנית‬technical language in rabbinic literature), Lešonenu 26 (1962), p. 215 [Hebrew]; H. Rabin, “‫ שנה‬125 ‫ =( ”עברית מדוברת לפני‬Spoken Hebrew 125 years ago), Leshonenu Laʿam 14 (1963), issue 5 [137], pp. 116-117 fn. 17; Sivan 1966, p. 191. The term batte ʿenayim also appears in a 1904 issue of BenYehuda’s own newspaper the Hashkafa, see Ornan 1996, p. 47. 58 Renassia 1930a, p. 225: généralissime – ‫( שַ ר ַצבָא‬sic!) (sare tsaba) – ‫( ﻗﺎﻳﯾﺪ ﺟﻴﯿﺶ‬k’aide djiche). Also see Sivan 1966, p. 192. 59 Renassia 1930a, p. 50: boussole – ‫( ַמנְהִיג ַה ָמ ָלחִים‬sic!) (manehigh hammallah’ime) – ‫( ﻣﺎﻋﻮﻥن ﻳﯾﻮﺭرﻱي ﺍاﻟﻄﺮﻳﯾﻖ‬maône iouerri etrik). Also see Sivan 1966, p. 190. 60 Renassia 1930a, p. 7: aéroplane – ‫( ְמעֺ ֵפפָה‬méôphépha) – ‫( ﻗﺒﺔ ﺍاﻟﻬﮭﻮﺍاء‬kobbeth lahoua); p. 32: ballon – ‫( ְמעֺפֵף‬méôphépha) – ‫( ﻁطﻴﯿﺎﺭرﺓة‬tiyara) [note that ‫ מְ עֺפֵף‬is masculine but its transcription méôphépha is feminine]. Cf. Renassia 1930a, p. 7: aérostat – ‫אֺהֶל‬ ‫פוֹר ַח‬ ֵ ‫ ִמגְדָ ל‬/‫( זָרוֹק‬ohèle zarok’/mighdal poreyah’) – ‫( ﻁطﻴﯿﺎﺭرﺓة‬tyara). The term migdal haporeaḥ ba-ʾavir (‫ )מגדל הפורח באויר‬for ‘a balloon’ is found in the writings of the nineteenth century Jewish Enlightment scholar Z. H. Slonimski, see Kaddari 2002, p. 39. This use is based on the occurrence of this term in the Talmud (b. Ḥagiga 15b, b. Sanhedrin 106b) albeit in a different meaning (an open air tower). 61 Ben-Yehuda 1948, vol. 6, pp. 3156-3157; Ornan 1996, p. 177. 62 The word ʾaviron (‫ = ;אוירון‬airplane) does not appear in Ben-Yehuda’s dictionary. Under the entry ʾavir (‫ = ;אויר‬air) he lists only sәfinat ʾavir (‫ = ;ספינת אויר‬airship) and sappan ʾavir (‫ = ;ספן אויר‬airship pilot). See Ben-Yehuda 1948, vol. 6, pp. 3156-3157. Reʾuven Sivan comments: ‫ =( המלה אוירון הובאה רק בין שייכי ערך מלחמה ללא הגדרה‬the word ʾaviron was listed among ‘war’ entries without a definition), see R. Sivan, ‫"חידושי‬ "‫ =( המלים של אליעזר בן יהודה לפי מילונו‬Eliezer Ben-Yehuda’s word innovations according to his dictionary), Leshonenu Laʿam 12 (1961), issue 2-3 [114-115], p. 43. Also see Ornan 1996, pp. 6 (‫)אוירון‬, 177 (‫ ;)מעופף‬Sivan 1966, p. 217 (‫[ )אווירון – מטוס‬Hebrew]; R. Sivan, "‫ אוירון‬.‫ קכ‬:‫ =( "מחיי מלים‬From the life of words: 120. ʾaviron), Leshonenu Laʿam 23 (1972), issue 9 [229], pp. 254-255.

Ofra Tirosh-Becker

444

Interestingly, both ʾaviron and mәʿofef were pushed aside by the noun maṭos (‫ )מָטוֹס‬and the verb ṭas (‫ ) ָטס‬that were later proposed by H. N. Bialik, who advocated the use of a single root for all verbs and nouns related to flight, such as maṭos (‫ ;מָטוֹס‬plane), ṭayyas (‫ ; ַטיּ ָס‬pilot); ṭayis (‫ ; ַטי ִס‬flying); ṭisa (‫ ;טִיסָה‬flight), and ṭayyeset (‫ ; ַטיֶּסֶת‬squadron).63 Another example: for the French word parapluie, which means ‘an umbrella’, the Algerian dictionary gives the Hebrew maṭṭara (‫ ; ַמטּ ָָרה‬which in contemporary Hebrew is used for ‘a target’) with an alternative rendering maḥase (‫ ) ַמ ֲחסֶה‬that denotes ‘a shelter’.64 The word maṭṭara shares its root √mṭr (that is related to ‘rain’) with the word miṭriyya (‫) ִמט ְִריּ ָה‬, which was coined by BenYehuda for this object. In this case Ben-Yehuda used the known Arabic suffix -iyya while Renassia preferred the Hebrew suffix -a.65 Among Ben-Yehuda’s innovations, which were either not known in Algeria or not adopted by the author of this dictionary, are words such as ʾadišut (‫ ;אֲדִ ישׁוּת‬indifference) and ʿagvaniyya (‫ ; ַעגְ ָבנִיּ ָה‬tomato).66 When Ben-Yehuda encountered missing lexemes that were necessary for rejuvenating the Hebrew speech he first and foremost relied on earlier strata of Hebrew and on Aramaic. However, when he did not find a suitable solution there, he turned to Arabic looking for appropriate roots.67 Examples are words such as hagira (‫ִירה‬ ָ ‫ ) ֲהג‬for ‘immigration’ based on the Arabic root √hǧr (‫)ﻫﮬﮪھﺠﺮ‬, rišmi (‫שׁמִי‬ ְ ‫)ר‬ ִ for ‘official’ based on Arabic rasmi (‫)ﺭرﺳﻤﻲ‬, bәsima (‫שׂי ָמה‬ ִ ‫ ) ְבּ‬for ‘a smile’ re68 flecting the Arabic basma, ʾibtisāma (‫ ﺍاﺑﺘﺴﺎﻣﺔ‬,‫)ﺑﺴﻤﺔ‬, ʾadiv (‫ )אָדִ יב‬for ‘polite’ based on ָ ‫ )מִדְ ָר‬for ‘school’ based on Arabic madrasa Arabic ʾadib (‫)ﺃأﺩدﻳﯾﺐ‬,69 and midraša (‫שה‬ 63 Sivan 1966, pp. 215-216; S. Barak and R. Gadish (eds.), Safa Qama: Selections from the Leshonenu Laʿam Column, Haʾaretz, 1932-1944, Jerusalem: The Academy of the Hebrew Language 2009, p. 158, §325 [Hebrew]. 64 Renassia 1930a, p. 323: parapluie – (‫( ַמטּ ַָרה )מחסה‬sic!) (mattara) – ‫( ﺳﻴﯿﻮﺍاﻧﺔ‬siouana). 65 Felman 1989-1990, pp. 215-221 (p. 219: ‫)מטריה‬. D. Almagor, "‫"פיצוחייה – יה – יה‬, Leshonenu Laʿam 45 (1994), issue 2, p. 55 [Hebrew]. 66 On ʿagvaniyya (‫ ) ַעגְ ָבנִיּ ָה‬see R. Sivan, “‫ =( ”העגבנייה ומה שעוללו לה שמותיה‬The ʿagvaniyya and its names), Leshonenu Laʿam 22 (1971), issue 3 [213], pp. 77-104 [Hebrew]. The words rišmi (‫שׁ ִמי‬ ְ ‫)ר‬ ִ and bәsima (‫שׂי ָמה‬ ִ ‫ ) ְבּ‬discussed below are additional examples for Ben-Yehuda’s innovations that do not appear in this Algerian dictionary. 67 Felman 1989-1990, p. 219. 68 Ornan 1996, pp. 46-47. 69 Ben-Yehuda suggests that there may be an ancient Hebrew origin for this word as well, see Ben-Yehuda 1948, vol. 1, p. 59 (entry ‫)אדיב‬, fn. 1.

Eliʿezer Ben-Yehuda and Algerian Jews: Relationship and language

445

(‫)ﻣﺪﺭرﺳﺔ‬.70 We identify this very same logic – possibly influenced by BenYehuda’s approach – in Renassia’s trilingual dictionary. At times this Algerian author turns to Arabic for solving lexical gaps in Hebrew. In fact, Renassia was very much aware of the cognate nature of Hebrew and Arabic and has published multiple cognate pairs in his book Analogies hébraico-arabe (Constantine, circ. 1930).71 In some cases Renassia includes terms in his dictionary that are most likely calques of Arabic counterparts. Thus, for ‘barometer’ (F. baromètre) Renassia brings the term mozne ʾavir (‫ ;מֺאזְנֵי הָאַ ִויר‬air balances),72 which is similar to BenYehuda’s Arabic inspired mozen ʾavir (‫ ;מֺאזֶן האויר‬air balance),73 although we cannot know whether in this case Renassia did not directly rely on the Arabic counterpart mīzān ǝl-hawāʾ (‫)ﻣﻴﯿﺰﺍاﻥن ﺍاﻟﻬﮭﻮﺍاء‬, which was used in Algeria as well.74 Another example is the Hebrew term ḥag ha-šoṭim (‫ ;חַג הַשּוֹטִים‬literally ‘the holiday of fools’) for ‘carnival’ based on the Maghrebi Arabic term ʿid al-mahābil (‫)ﻋﻴﯿﺪ ﺍاﻟﻤﻬﮭﺎﺑﻞ‬.75 Also note his use of ṣәvaʾi (‫ ) ְצ ָבאִי‬for ‘a soldier’ (F. soldat), which reflects the Arabic word ʿaskari (‫)ﻋﺴﻜﺮﻱي‬,76 not using the maskilim’s term ʾiš ṣava (‫ ) ִאיש ָצבָא‬nor Ben-Yehuda’s innovation ḥayyal (‫) ַחיּ ָל‬.77 Likewise, Renassia suggested ṣura (‫)צוּרה‬ ָ for ‘a photo’ (a noun that is used in Modern Hebrew for ‘shape, form’), clearly mirroring the Arabic term ṣūra (‫)ﺻﻮﺭرﺓة‬,78 again not men-

70 See Ben-Yehuda 1948, vol. 6, p. 2818 (entry ‫ )מדרשה‬where he explains: ‫שם לבית ספר‬ ‫ נהוג בדיבור העברי בא"י והשתמשו בו בהעתונים‬,‫ בין בית ספר למתחילים ובין בתי הספר העליונים‬,‫תיכון‬ (= A name for a middle school, between a beginners’ school and high schools, it is commonly used in Hebrew speech in the Land of Israel and in the journals (HaʾOr, Ḥeshvan 1914). In a footnote he adds: .‫ =( כך בערב' מדרשׂה‬Cf. in Arabic madrasa). 71 Renassia 1930b. 72 Renassia 1930a, p. 34: baromètre – ‫( מֺאזְנֵי הָאַוִיר‬sic!) (mozné-haavire) – ‫ﻣﻴﯿﺰﺍاﻥن ﺍاﻟﻬﮭﻮﺍاء‬ (mizane lahoua). 73 Ben-Yehuda 1948, vol. 1, p. 97 (in the entry ‫)אויר‬: ‫=( מאזן האויר – כלי לשקל את לחץ האויר‬ an instrument to weigh air pressure). 74 Beaussier 1958, p. 1054. 75 Renassia 1930a, p. 66: carnaval – ‫( חַג הַשוֹטִים‬h’agh hachotim) – ‫( ﻋﻴﯿﺪ ﺍاﻟﻤﻬﮭﺎﺑﻞ‬îde elemehabèl). 76 Renassia 1930a, p. 413: soldat – ‫( ְצ ָב ִאי‬tsébaï) – ‫( ﻋﺴﻜﺮﻱي‬âskri). 77 Sivan 1966, p. 191; R. Sivan 1973, pp. 91-92. 78 Renassia 1930a, p. 337: photo – ‫צוּרה‬ ָ (tsoura) – ‫( ﺻﻮﺭرﺓة‬tsoura); p. 213: forme – ‫צוּרה‬ ָ (tsoura) – ‫( ﺻﻮﺭرﺓة‬tsoura),

446

Ofra Tirosh-Becker

tioning the older two-word term ṣiyyur ʾor (‫)צִיּוּר אוֹר‬79 nor David Yellin’s innovation ṣillum (‫)צִלּוּם‬.80 5. Conclusion To conclude, in this paper we shed light on several aspects of the relationship between Ben-Yehuda and Algerian Jews. First, we discussed Ben-Yehuda’s sojourn in Algiers early in his life, and the impact that his exposure to Hebrew speech there had made on him and on his future preference for the Sephardic pronunciation. Next we accounted for the high stature of Ben-Yehuda in the eyes of Algerian Enlightment scholars, who revered him as one of the cornerstones of Zionism. Finally, we presented the impact that Hebrew revival, championed by Ben-Yehuda and his colleagues in the Land of Israel, had made on Hebrew in Algiers as exemplified by their language innovations that found their way into a tri-lingual French-Hebrew-Arabic dictionary composed in Constantine around 1930.

References81 Beaussier 1958 = Marcelin Beaussier, Dictionnaire pratique arabe-français: contenant tous les mots employés dans l’arabe parlé en Algérie et en Tunisie, Nouvelle édition, revue, corrigée et augmentée par M. Mohamed ben Cheneb, Alger 1958. Ben Sedira 1995 = Belkassem Ben Sedira, Dictionnaire Arabe-Français (Langue, Letters, Conversation), Beirut 1995. Ben-Yehuda 1948 = Eliʿezer Ben-Yehuda, A Complete Dictionary of Ancient and Modern Hebrew, Jerusalem 1948. [Hebrew] Chetrit 1990 = Joseph Chetrit, “Hebrew National Modernity against French Modernity: The Hebrew Haskalah in North Africa at the end of the Nineteenth Century”, Miqqedem Umiyyam 3, Tradition and Modernity in the North African and Oriental Jewry (1990), pp. 11-76. [Hebrew] Chetrit 1993 = Joseph Chetrit, “Changes in the Discourse and Arabic Language of the Jews of North Africa at the end of the Nineteenth Century”, Peʿamim 53 (1993), pp. 90-123. [Hebrew]

79 Ornan 1996, p. 260. 80 Sivan 1966, p. 190. See Renassia 1930b, p. 122: forme – ‫ – צורה‬ṣūra – tsora. 81 Complete bibliographic details of references that are cited in this paper only once are given in the footnotes. The list below includes only references that were cited more than once using the abbreviations.

Eliʿezer Ben-Yehuda and Algerian Jews: Relationship and language

447

Eldar 2010 = Ilan Eldar, Language Planning in Israel, Jerusalem: The Academy of Hebrew Language, 2010. [Hebrew] EJ2 = Encyclopaedia Judaica, eds. Michael Berenbaum and Fred Skolnik, 2nd edition, Vol. 15. Detroit: Macmillan Reference USA 2007, pp. 297-299. EJIW = Encyclopedia of Jews in the Islamic World, editor-in-chief Noam A. Stillman, Leiden: Brill Academic Publishers 2010; and see 2nd edition online. Felman 1989-1990 = Yaacov Felman, "‫ =( "אליעזר בן יהודה ותחיית הלשון העברית‬Eliezer BenYehuda and the revival of the Hebrew language), Leshonenu Laʿam 40-41 (19891990), pp. 215-221. [Hebrew] Harshav 1990 = Benjamin Harshav, “‫ =( ”מסה על תחיית הלשון העברית‬a thesis on the revival of the Hebrew language), Alpayim – A Multidisciplinary Publication for Contemporary Thought and Literature 2 (1990), pp. 9-54. [Hebrew] Hava 1970 = Joseph G. Hava, Al-Farâʾid Arabic-English Dictionary, Beirut 1970. Kaddari 2002 = Dorit Kaddari, Nominal Phrases in Prose from the Haskala Period, Master’s thesis, The Hebrew University of Jerusalem, 2002. [Hebrew] Marciano 2002 = Eliahou Marciano, Les sages d'Algérie: Dictionnaire encyclopédique des sages et rabbins d'Algérie, du haut Moyen-âge à nos jours, Marseille: IMMAJ 2002. Morag 1957 = Shlomo Morag, “ ?‫ עד מתי דיברו עברית‬.‫ א‬:‫“ =( ”לשון בגלותה ובתחייתה‬A language in exile and in its revival: a. Till when was Hebrew spoken?”), Leshonenu Laʿam 7 (1957), issue 5-6 [67*-68*], pp. 3-10. [Hebrew] Ornan 1996 = Uzzi Ornan, The Words Not Taken: A Dictionary of Forgotten Words, Jerusalem: The Magnes Press 1996, p. 177. [Hebrew] Renassia 1930a = Joseph Renassia, Dictionnaire Français–Hébreu-Arabe, Constantine: Imp. Attali, 1930? Renassia 1930b = Joseph Renassia, Analogies hébraico-arabes: mots arabes tirés du Pentateuque, des Prophétes, des Hagiographes et du Talmud, Constantine: Imp. M. Attali, circ. 1930. Reshef 2014 = Yael Reshef, " ‫"אליעזר בן יהודה ובני דורו‬, in Nitʿe Ilan: Studies in Hebrew and Related Fields Presented to Ilan Eldar, Moshe Bar-Asher and Irit Meir (eds.), Jerusalem 2014, pp. 613-624. [Hebrew] Sivan 1966 = Reuven Sivan, "‫ שיחות על חידושי מלים‬:‫ =( "לשון בתחייתה‬Language in its revival: Conversations on words innovation), Leshonenu Laʿam 17 (1966), issue 7-8 [169170], pp. 177-234. [Hebrew] Sivan 1973 = Reuven Sivan, "‫ =( "אליעזר בן יהודה ותחיית הדיבור העברי‬Eliezer Ben-Yehuda and the Revival of Hebrew Speech: Chapter 3: ‘Words tell’), Leshonenu Laʿam 24 (1973), issue 3-4 [233-234], pp. 67-116. [Hebrew] Sivan 1988 = Reuven Sivan, "‫ חידושי לשון‬.‫ ב‬:‫ =( "עוד הוויות וחויות בלשון‬More on language and its experience: Language innovations), Leshonenu Laʿam 39 (1988), issue 1-2 [381382], pp. 6-47. [Hebrew] Tirosh-Becker 2011 = Ofra Tirosh-Becker, “The Languages of Algerian Jewry”, in Algeria: Jewish Communities in the East in the Nineteenth and Twentieth Centuries, ed. Haim Saadoun, Jerusalem: Ben-Zvi Institute 2011, pp. 117-132. [Hebrew]

Explicit performative utterances in Semitic Rebecca Hasselbach-Andee, University of Chicago

1 Introduction1 Performative utterances (short: performatives) are roughly defined as “a type of sentence where an action is ‘performed’ by virtue of the sentence having been uttered”.2 These sentence types are commonly distinguished from descriptive statements, as in the examples ‘I will come to you tomorrow’ and ‘I came to you yesterday’ (descriptive) versus ‘I promise you to come to you tomorrow’ (performative, act of promise being performed). Performatives are a cross-linguistically well-known phenomenon and it is thus no surprise that they are also a common feature in Semitic languages – at least in those Semitic languages for which we have sufficient evidence to identify them. The existence of performatives in Semitic languages has already been noted in early descriptions of individual Semitic languages. Although usually called differently, reference grammars such as Gesenius and Kautzsch (Biblical Hebrew), Wright (Classical Arabic), Dillmann and Bezold (Classical Ethiopic), and others cite undisputable examples of performative utterances.3 As described in more detail in section 3 below, cross-linguistically many languages use a form of the present tense to express performatives. Given the strong present-tense association of performatives due to their very nature as “performing” an act while being uttered, this is not surprising. A few languages, including many Semitic languages, however, use non-present tense forms. 1

The following abbreviations are used in the glosses: ACC = accusative; ASSEV = asseverative particle; AUX = auxiliary; COMP = complement; CSTR = construct; DAT = dative; DN = divine name; FUT = future; GEN = genitive; GN = geographical name; IMP = imperative; IMPF = imperfect; INDEP = independent pronoun; INFIN = infinitive; PART = particle; PERF = perfect; PL = plural; PN = personal name; PREP = preposition; PRET = preterite; PTC = participle; REL = relative particle; SING = singular; VENT = ventive; VOC = vocative.

2

Crystal 2003: 343.

3

Gesenius and Kautzsch 1910: 311-312; Wright 1898 II: 1; Dillmann and Bezold 1907: 168.

Explicit performative utterances in Semitic

449

In Semitic (see section 2 below), it is usually a past-tense form, such as the preterite or perfect in Akkadian and the perfect in West Semitic languages that expresses performatives. Given the aforementioned present-tense association of performatives, the use of a verbal form whose primary function is to express past tense/perfective aspect requires explanation. Several suggestions have been made as to why Semitic languages exhibit this seemingly unusual feature. The most common explanation is the assumption that performatives are viewed as completed when uttered and thus the perfect(ive) is used in order to reflect the notion of completion. This analysis of performatives in Semitic is connected to an aspectual understanding of the respective verbal forms.4 The interpretation of performative utterances as reflecting completed actions, however, faces several problems. First of all, it is based on an outdated definition of perfective aspect as expressing completed actions/situations. Second, it is questionable if performative utterances, which inherently refer to the present and are often referred to as prototypical present-tense situations because they are performed while being spoken, can truly be understood as “completed”. Because of these problems it is worthwhile to look for a different explanation for the use of the preterite/perfect for performatives in Semitic, which is the aim of this study. The following investigation starts out with a general discussion of performative utterances. It then moves on to provide an overview of the evidence from individual Semitic languages, previous literature on and explanations of performatives in Semitic, and ultimately leads to an alternative explanation for the use of verbal forms in these types of utterances. 1.1 Performatives In “How to do Things with Words” (1962), J.L. Austin defined the basic notion of performative utterances and laid the foundation for contemporary studies of these types of sentences in English-speaking scholarship. As is well known, the German Slavicist E. Koschmieder previously developed a similar theory (e.g. Koschmieder 1945), in which he called the phenomenon “Koinzidenzfall”. Koschmieder’s approach has widely been followed in German scholarship on Semitic languages. In their basic essence, performative and “Koinzidenzfall” describe the same phenomenon, namely a speech type in which the utterance of a sentence performs the act that the verb denotes. Both Koschmieder and 4

See, e.g., Koschmieder 1945: 28; Weninger 2001: 78; Gzella 2004: 214. For a more detailed discussion and references see section 3 below.

450

Rebecca Hasselbach-Andee

Austin further note that the verb in such utterances usually appears in the first person.5 Commonly cited examples of performatives include “I promise that…”, “I state that…”, “I warn you that…”, “I apologize”, etc. In English, the particle “hereby” can be inserted into such utterances, as in “I hereby promise…” etc. The insertion of “hereby” clearly marks the sentence as performative and distinguishes it from other possible readings such as the progressive present and habitual acts.6 Since the utterance itself performs the act, according to Austin, it does not have a truth factor or truth condition. This means that performatives cannot be true or false unlike descriptive statements.7 Compare the descriptive sentences “yesterday I went to the store” and “I will go to the store tomorrow”, which, depending on circumstances, describe a true event, that is, an event that did or will happen or a false event that did not or will not occur, with the performative “I promise I will go to the store”. The latter cannot be true or false because the act of promise is performed by uttering the sentence itself.8 Austin thus distinguishes between descriptive statements that can be true or false, which he calls “constative”, and utterances in which something is done by saying it and that are lacking a truth factor, which he calls “performative”.9

5

Koschmieder 1945: 23; see also Recanati 1987: 45-46.

6

See, e.g., Rogland 2001: 243.

7

Austin 1962: 5.

8

It is, of course, possible to make a false promise, that is, I promise to do something but have no intention of keeping the promise or there are circumstances that prevent me from keeping my promise. Austin, however, distinguishes this type of “false promises” from the truth factor of descriptive statements. In performatives, the act of promise etc. is performed no matter what the intentions of the speaker might be. This differs from descriptive sentences, in which the very statement is subject to being either true or false. Austin thus uses a different term regarding situations in which a performative can go wrong: “felicity” (Austin 1962: 14). A performative that fulfills all conditions for being successful, that is, I make a promise, I intend to keep it, and all conditions are met for me to keep it, is “happy” or “felicitous”. A performative in which not all necessary conditions are met, on the other hand, is called “unhappy” or “infelicitous” (see Austin 1962: 14-22 for various situations in which a performative can be “unhappy”). See also Levinson 1983: 229 for Austin’s “infelicitous” categories.

9

Austin 1962: 3, 13; see also Recanati 1987: 68. Koschmieder provides the same basic definition of what he calls “Koinzidenzfall”. According to Koschmieder, “Koinzidenzfall” means that the action happens at the same time (thus “Koinzidenz”) as the utterance. This means that the utterance is the action itself (Koschmieder 1945:

Explicit performative utterances in Semitic

451

This basic constative/performative distinction, however, only reflects the initial stage of Austin’s analysis, which he further developed from the basic constative/performative distinction to a more fully developed speech act theory. Austin’s speech act theory still includes performative utterances of the type just cited, that is, utterances characterized by the type of verb they employ (such as ‘to promise’, ‘to state’, ‘to bet’, etc.), the fact that the verb most often occurs in the first person, and the possibility to insert particles such as ‘hereby’ into the sentence. Austin calls this type of utterance “explicit performative”. Austin also notes, however, that there are many more types of performative utterances than explicit performatives, where the primary mode of expression does not lie with a performative verb. Performatives can also be expressed by mood, stress, adverbs, adverbial expressions, conjunctions, etc. Performatives that do not include a performative verb are called “primary” or “implicit performatives”.10 An example would be a sign posted on the fence of a meadow saying “dangerous bull!”, which implies a warning. According to Austin, the important test for implicit performatives is that it must be possible to reformulate them as explicit performatives, as in “I hereby warn you of the dangerous bull!”.11 Most importantly, Austin’s speech act theory distinguishes between 1. a locutionary act/force, which describes the actual utterance, that is, the act of “saying”, and 2. an illocutionary act/force, which denotes the intention of the speaker or pragmatics of using a locutionary act. Widely accepted illocutionary acts are those of warning, promising, requesting, baptizing, describing, etc.12 All speech acts, whether explicit or implicit, are commonly understood as expressing illocution. In other words, illocution is equated with the notion of speech acts or performativity itself.13 Locution and illocution are thus not mutually exclusive concepts. In fact, they represent simultaneous aspects of an utterance. It is impossible to perform a locutionary act without performing illocution at the 22). In German, these types of utterances can be distinguished from circumstantial clauses by inserting the particle “hiermit” (ibid.). 10 See the descriptions in Levinson 1983: 231 and Wagner 1997: 9. 11 Wagner 1997: 10. 12 Wagner 1997: 11-12. A third act/force is the perlocutionary force, in which “the act is defined by reference to the effect it has on the hearer”, as in cases of convincing, deterring, persuading, frightening, etc. (Crystal 2003: 226; Wagner 1997: 11-12). The difference between illocutionary and perlocutionary acts are that the former are not defined with reference to the hearer (see Levinson 1983: 236). 13 Crystal 2003: 226.

452

Rebecca Hasselbach-Andee

same time. This also means that Austin’s original distinction between constative/ performative is no longer valid because constative acts such as describing, ascertaining, claiming, etc., are now considered categories of illocution: both performatives and constatives are illocutionary acts.14 Based on this argument, Austin concludes that there are no true features that distinguish performative from constative utterances since all utterances “do something”.15 Speech act theory consequently makes finer distinctions in terms of illocution than the initial dichotomy of performative/constative. In fact, it gets rid of the performative/constative distinction altogether based on the definition of both utterance types as reflecting illocution, although different types of illocution.16 Based on speech act theory, explicit performatives serve to perform the illocutionary act denoted in the performative verb (e.g. ‘I order that..’, ‘I promise that…’). In other words, in explicit performatives, illocution is expressed by an explicitly stated performative verb that is followed by a proposition in the rest of the utterance, as in “I hereby promise I will come tomorrow”.17 In the study of Semitic languages, especially in German scholarship in which Koschmieder’s influence was quite strong, the investigation of per14 Wagner 1997: 12; Recanati 1987: 171. 15 See description in Wagner 1997: 11. 16 Wagner 1997: 13. For a nice discussion of the development from the basic constative/performative distinction to speech act theory see Wagner 1997: 9-13. Austin’s speech act theory has subsequently been expanded and modified, see e.g. the description of Searle’s approach in Wagner 1997: 14-15. Searle places more focus on the context of speech acts. The context or rules are associated with conventions and institutions – “institutions” in this case means a system of rules, the behavior etc. that reflects a specific culture (ibid.). For the importance of context, that is, conventions, for performatives see also the discussion in Recanati 1987: 75-77. Recanati cites Austin’s interpretation of explicit performatives as representing conventional acts, as in the classic examples “I hereby declare you man and wife”, for which certain extra-linguistic conditions have to be met in order for the performative to be “felicitous”. Conventions are thus extralinguistic in nature (ibid.). Such conventional performatives cease to exist when the appropriate institution is absent. Not all explicit performatives, however, require extralinguistic context. Utterances such as “I state that..”, “I advise you…”, “I command you…” are speech acts without such explicit contexts, although also utterances such as these require a certain conventionalism in the sense that I can only command an inferior, not a superior (Recanati 1987: 77-80). Koschmieder’s analysis never developed further than the basic definition of of performatives (Wagner 1997: 56). 17 Wagner 1997: 44-45.

Explicit performative utterances in Semitic

453

formatives is mostly limited to the study of performative verbal constructions in the form of explicit performatives.18 Much less often, studies e.g. of the Hebrew verbal system use the concept of speech act theory.19 Explicit performatives, as Wagner states, reflect the most distinct category of performatives, but they only constitute a small part of the overall speech act system.20 Despite the fact that I am aware of the limited scope that explicit performatives represent in terms of types of speech acts in general, this study nevertheless focuses on explicit performatives since its aim is to analyze the use of verbal forms in performative utterances in Semitic. The reader should, however, be aware that this category of speech acts only represents a sub-category of a much more extensive phenomenon. 1.2 Explicit performatives and epistolary perfects in Semitic Before we can discuss explicit performatives in Semitic, an important distinction has to be made between two similarly expressed but functionally quite different constructions that have often been conflated in studies of Semitic languages: explicit performatives and epistolary perfects (expressed by the preterite and perfect in Akkadian).21

18 Studies influenced by Koschmieder include e.g. Mayer 1976: 181-198; Weninger 2000: 91-99 and Weninger 2001; Gzella 2004: 205-215. For a general discussion of Austin’s and Koschmieder’s approaches and their application to Semitic see Wagner 1997: 59-67. For studies on performatives that focus on the verbal expression of such utterances, see e.g. Heimpel and Guidi 1969, Hillers 1995, Manahlot 1988, Mayer 1976, Pardee and Whiting 1987, Rogland 1999, Rogland 2001, and Weninger 2000. 19 See, e.g. Wagner 1997. 20 Wagner 1997: 75. Wagner stresses that is it important to look at other types of performatives and their expression, such as lexical and syntactic ways to code speech acts (1997: 77). In his study, Wagner investigates various types of speech acts in Biblical Hebrew in order to find the indicators of illocutionary force in this language. His study is one of the most comprehensive ones on performatives in a Semitic language. 21 Heimpel and Guidi argue that the preterite and perfect serve different functions in Akkadian letters: the preterite of the verb šapārum ‘to send’ can be performative while the perfect seems to reflect epistolary usage (Heimpel and Guidi 1969: 151). This interpretation has rightly been rejected since both preterite and perfect can be used for “epistolary perfects” (see, e.g. Pardee and Whiting 1987: 27; Rogland 2001: 247). The difference rather seems a diachronic one (see section 2.1 below for the

454

Rebecca Hasselbach-Andee

Epistolary perfects, as the name indicates, occur in letters and often indicate actions such as sending, sealing, bowing down, etc., as in: Akkadian:22 ina qabê bēlī-ya ašpur-ak-kum in command.CS lord.GEN-1CS.GEN 1CS.PRET.send-VENT-2MS.DAT ‘by the command of my lord, I am (hereby) writing to you…’ (RA 21,42,11) ana PN qibī-ma umma PN aššum DN ana GN alākim to GN go.INFIN to PN say.IMP.MS-PART PART PN because DN ṭupp-ī ašpur-am 1CS.PRET.send-VENT tablet-1CS.GEN ‘to PN say, thus PN: I am (hereby) writing concerning the journey of DN to GN’ (AbB 2,63) Hebrew: šālaḥ-tî lә-kā šōḥad send.PERF-1CS to-2MS gift ‘I am (hereby) sending you a gift’ (1 Kgs 15:19)

ʾîš ḥākām send.PERF-1CS man wise ‘I am (hereby) sending a craftsman’ (2 Chr 2:12) šālaḥ-tî

Imperial Aramaic:23 l-šlm-ky šlḥt sprh znh to-well.being-2FS send.PERF-1CS letter DEM ‘I am (hereby) sending you this letter for your well-being’ hwšr-t l-k w-šrrt šgyʾ and-strength abundant send.PERF-1CS to-2MS ‘I (hereby) send you abundant welfare and strength’ šlm welfare

substitution of the preterite by the perfect in main clauses from the mid-second millennium BCE on). 22 See Heimpel and Guidi 1969: 151-152. 23 See Hug 1993: 116; Muraoka and Porten 2003: 193-194.

Explicit performative utterances in Semitic

455

In Semitic languages, epistolary perfects are usually expressed by the same verbal forms as explicit performatives (for which see section 2 below). Given the fact that the context of epistolary perfects often also allows the insertion of ‘hereby’, as seen in the examples cited above, they have often been interpreted as performatives or have simply been listed together with true performatives in reference grammars and/or studies of the use of verbal forms in individual Semitic languages.24 The seeming similarities of epistolary perfects and explicit performatives, however, are misleading since the two constructions have different functions. The main difference between epistolary perfects and performatives is that epistolary perfects do not perform an act as explicit performatives do, they simply report an act, such as the act of sending in the examples given above. In these cases, the act of sending cannot be realized through the utterance alone, as opposed to utterances such as ‘I (hereby) bless you/promise you’, etc.25 Epistolary perfects thus have the illocutionary force of description and differ in this regard from explicit performatives, despite the formal similarities between the two constructions in Semitic languages, which still require explanation. The most commonly found explanation is that the writer of the letter uses the perfect/preterite because he writes from the perspective of the reader. For the reader, the act of sending, writing, sealing, etc. has already happened, and thus a past-tense form is used.26 This kind of construction is comparable to English phrases such as “I have enclosed a copy of the lecture” when writing a letter, which likewise uses a past tense verbal form because the letter writer assumes the viewpoint of the reader.27 The present study follows this analysis of epistolary perfects.28 24 See, e.g., Hug 1993: 116; Joüon and Muraoka 1996: 362; Muraoka and Porten 2003: 193-194, to name just a few. 25 Pardee and Whiting 1987: 25; Streck 1995: 158. 26 Streck 1995: 165; Rogland 2001: 247-248. See also Comrie 1985: 16 for the fact that we sometimes find the use of the recipient’s deictic center, that is, past tense for the writer’s present moment. Instances of past-tense forms for these types of contexts are not uncommon cross-linguistically. They are, for example, well attested in Latin and Greek (Pardee and Whiting 1987: 3). 27 Rogland 2001: 248. 28 Rogland suggests that the most fitting translation for epistolary perfects would be a past tense verb, such as “I have sent…” instead of a present tense verb as given in many reference grammars and in the examples above (Rogland 2001: 248). The choice of translation value depends on which perspective the translator wishes to convey: that of the letter writer or the reader.

456

Rebecca Hasselbach-Andee

Important for the context of this study is that epistolary perfects reflect tense-use of the verbal form in question: the perfect/preterite is used because from the viewpoint of the reader, the action described has happened in the past. As argued below (section 3.1), the motivation for expressing explicit performatives with the perfect/preterite differs significantly from epistolary perfects. The two constructions, although superficially similar, thus have to be clearly distinguished from each other. Epistolary perfects will not be considered any further in this study. 2 Semitic evidence for explicit performatives Performatives are, as mentioned in the introduction, attested in almost every Semitic language, although it can be difficult to identify this type of utterance in purely epigraphic languages. The following examples have been chosen to represent widely accepted cases of explicit performatives. The languages used for the present investigation include Akkadian, Biblical Hebrew, Ugaritic, Classical and Syrian Arabic, Ethiopian Semitic, Aramaic (various dialects), and Mehri. Wherever possible, changes in the expression of explicit performatives throughout the history of a language have been noted. 2.1 Explicit performatives in Akkadian Explicit performatives are attested in almost all dialects of Akkadian. Already von Soden noted that explicit performatives in Old Babylonian are expressed by the preterite:29 umma šū-ma šarr-am a-tma ištu inanna from now thus 3ms.INDEP-PART king-ACC 1CS-swear.PRET adi ūm-im ḫamš-im kasap-ka lū a-naddik-kum silver.CSTR-2MS ASSEV 1CS-give.IMPF-2MS.DAT until day-GEN 5-GEN ‘Thus he: I (hereby) swear by the king that I will repay you your silver in five days from now’ (VAB 6, 207:21) a-lsī-ka 1CS-call_out.PRET-2MS.ACC ‘I (hereby) call out to you, [o God NN]’

29 Von Soden 1995: 129.

Explicit performative utterances in Semitic

457

The use of the preterite for explicit performatives is also, although rarely, attested in Old Akkadian, as in the following lines from a love incantation from Kish:30 Inanna u

Išḫara u-tammē-ki

DN

DN

and

1CS-swear.PRET-2FS.ACC

‘I (hereby) conjure you by Inanna and Išḫara’ (MAD 5 8:33-34) It is further found in two passages from Old Akkadian letters:31 naḥas-su u-ma (spelled ù-má) life.CSTR-3MS 1CS-swear.pret ‘I (hereby) swear by his (the king’s) life’ (Ad 3:9) naḥas Šarkališarrī u-ma (spelled ù-má) 1CS-swear.PRET life.CSTR PN ‘I (hereby) swear by the life of Šarkališarrī’ (Gir 19:28-29) Although these two passages from letters clearly represent explicit performatives, it is difficult to determine the nature of the underlying verbal form. Commentators and grammars either do not explicitly mention the form itself or take it as a durative from wamûm.32 Mayer, however, interprets the spelling ù-má as preterite, that is uma instead of umma.33 The orthography is ambiguous and the form can be interpreted as both preterite and durative. Given the overwhelming evidence for the use of the preterite for explicit performatives in Akkadian as opposed to pretty much the non-existence of the durative for this function, Mayer’s reading might indeed be the better choice. In this case, the Old Akkadian forms should be interpreted as preterites, not duratives, as indicated in the glosses of the examples given above.34

30 See Westenholz 1977 for the edition of the text. 31 See Kienast and Volk 1995 for the edition of these texts. 32 See, e.g., Hasselbach 2005: 218. 33 Mayer 1976: 193. 34 For the interpretation of ù-má as durative against Mayer see Streck 1995: 93.

458

Rebecca Hasselbach-Andee

An example from Old Assyrian, in which a certain Buzazu is summoned to Assur and his reply most likely constitutes an explicit performative, likewise attests to the use of the preterite. Buzazu replies to the summons:35 u-škaʾʾin 1CS-submit.PRET

ana

ṭupp-im

to

tablet-GEN

ša

āl-imKI

u

rābiṣ-im

REL

city-GEN

and

bailiff-GEN

ša

bēl-ī-ya lord-GEN-1CS ‘I (hereby) submit to the tablet of the city and the bailiff of my lord’ (EL 325:2021) REL

During the third and first half of the second millennium BCE, Akkadian thus uses the preterite for explicit performatives. From Middle Assyrian on, however, Akkadian starts to use the perfect for this type of construction, as in the following oath context:36 a-ttama 1CS-swear.PERF ‘I (hereby) swear’ This change in the use of the underlying verbal form is commonly explained by the fact that the perfect is taking over many of the functions of the preterite at this time.37 Kouwenberg, in his recent study on the Akkadian verb (2010) states that the perfect and preterite have the same temporal function of referring to past events from Middle Babylonian and Middle Assyrian on.38 The difference between the two forms at this point is syntactical: the perfect occurs as regular past tense in affirmative main clauses while the preterite is used in subordinate and negative clauses. This means that the perfect replaced the preterite in main clauses and took over all or most of its functions.39 Important for the present

35 For the interpretation of this passage as explicit performative see Heimpel and Guidi 1969: 149. 36 Von Soden 1995: 130. 37 Von Soden 1995: 130. 38 Kouwenberg 2010: 153. 39 For an explanation how this reanalysis of the perfect occurred see Kouwenberg 2010: 154.

Explicit performative utterances in Semitic

459

context is the fact that, although the verbal form used for explicit performatives changed within the history of Akkadian, the underlying function of the verbal form remained the same, that is, the perfect serves the same function in this context as the original preterite. First millennium Akkadian reflects the same change in verbal forms, that is, the perfect is the most frequently attested form for explicit performatives, as in the following examples from Neo Babylonian and Neo Assyrian:40 ina libbi ilān-ī-ya a-ttama gods-GEN-1CS 1CS-swear.PERF by heart.CSTR ‘by my gods I swear’ (ABL 541 Rs 2) ina libbi by heart.CSTR

Aššur ilānē-ʾa

a-tteme

DN god-1CS 1CS-swear.PERF ‘by Aššur, my god, I swear’ (ABL 287, Rs 7)

Akkadian thus uses two main verbal forms for expressing explicit performatives: the preterite, which is attested in texts from the third and first half of the second millennium BCE, and the perfect, which starts to appear in the second half of the second millennium BCE and seems to be the primary form used in first millennium texts, based on currently available data.41 40 Mayer 1976: 194. 41 The ritual passages Mayer cites as examples of performatives that are expressed by the durative (Mayer 1976: 200) are not necessarily convincing. These texts rather reflect the performer of the ritual as describing present actions connected to the ritual, as in ‘I bring you (anaššī-kum) water’, ‘I dedicate to you (akarrab-kunūši)’ (ibid.). The contextual setting of these ritual passages thus differs from those encountered in oaths mentioned above. For criticism against Mayer’s interpretation of these types of iparras forms in rituals as performatives see also Streck 1995: 93. It is further debatable if the Late Babylonian example of a performative expressed by iparras that is quoted by Mayer is truly a performative. The form i-šá-al in PN šulum šá PN ŠEŠ-šú i-šá-al (TMH 2/3 260: 9-10), which Streck convincingly analyses as durative based on variant spellings such as i-šá-ʾ-al, can be translated as ‘PN (hereby) asks after the health of PN, his brother’. A minor issue is that the verbal form is in the third person, not the first person, although this can be explained by the fact that the letter-writer speaks of himself in the third person (Streck 1995: 93). The durative in this case could, however, also simply describe a present action from the perspective of the reader of the letter, that is ‘PN is asking after the health…’, which also fits the third person context better than a performative

460

Rebecca Hasselbach-Andee

2.2 Explicit performatives in Hebrew Biblical Hebrew contains numerous examples of explicit performatives.42 In Biblical Hebrew, these speech acts are expressed by the perfect, as in the following representative examples: nātat-tî ʾōt-әkā ʿal kol-ʾereṣ miṣrāyim rәʾē give.PERF-1CS ACC-2MS over all-land.CSTR Egypt see.IMP.MS ‘see, I (hereby) set you over all the land of Egypt’ (Gen 41:41) hēʾāsōp yē-ʾāsēp ʿālê-kā kî yāʿaṣ-tî for counsel.PERF-1CS be.gathered.INFIN 3MS-be.gathered.IMPF over-2MS kol-yiśrāʾēl all.CSTR-Israel ‘for I (hereby) counsel that all Israel be gathered to you…’ (2 Sam 17:11) h’ mәšaḥ-tî-kā lә-melek ʾel-yiśrāʾēl kō-ʾāmar thus-say.PERF.3MS Lord anoint.PERF-1CS-2MS to-king to-Israel ‘thus speaks the lord: I (hereby) anoint you king over Israel’ (2 Kings 9:3)43

ʾōt-ô hinne berak-tî behold bless.PERF-1CS ACC-3MS

wә-hiprê-tî

ʾōt-ô

and-make.fruitful.perf-1cs ACC-3MS ʾōt-ô wә-hirbê-tî and-make_numerous.PERF-1CS ACC-3MS ‘I (hereby) bless him and make him fruitful and numerous’ (Gen 17:20)44

interpretation. The use of the durative for performatives in Akkadian thus still requires further proof. 42 For a nice treatment of performatives in Biblical Hebrew and an extensive list of examples see Wagner 1997, in particular pp. 92-116. Performatives are further mentioned in various reference grammars of Biblical Hebrew. See, e.g., Gesenius and Kautzsch 1910: 311-312; Joüon and Muraoka 1996: 362; Waltke and O’Connor 1990: 488-489, although most of these cite explicit performatives together with epistolary perfects. 43 See also 2 Kings 9:6 and 2 Kings 9:12. 44 See also Gen 26:24, Ps 118:26, and Ps 129:8.

Explicit performative utterances in Semitic

way-yō-(ʾ)mer

b-î nišbaʿ-tî by-1cs swear.PERF-1CS

461

nәʾūm-h’ saying.CSTR-lord

and-3ms-say.IMPF ‘and he said: by myself I (hereby) swear, says the Lord’ (Gen 22:16)45 No examples seem to exist that are not expressed by the perfect in Biblical Hebrew. In modern Hebrew, explicit performatives are expressed by the participle, which has taken over the function of a present tense:46 Lapid: ani mavtiaḥ PN I promise.PTC

la-khem

she-kvar

tokh

to-2MP

that-already

within year

šana

va-ḥeṣi and-half

yi-hye yoter tov 3ms-be.IMPF more good ‘Lapid (minister of treasure): I promise you that already within a year and a half things will get better.’ (http://www.globes.co.il/news/article.aspx?did=1000852605) ani nišbaʿat

be-elohim hi naškh-a ot-i she bite.PERF-3FS ACC-1CS I swear.PTC by-God ‘I swear to god, she bit me’ (http://www.inn.co.il/News/News.aspx/259614) In Hebrew, it is thus possible to observe a shift in the verbal forms used for explicit performatives from perfect to participle.47

2.3 Explicit performatives in Aramaic Aramaic provides ample evidence for explicit performatives in its various dialects. The Old Aramaic texts treated in Hug’s grammar dating to the 7th and 6th centuries BCE attest to various cases of explicit performatives in letters and con-

45 See also 2 Sam 19:8. 46 I would like to sincerely thank Naʿama Pat-El for providing me with these modern Hebrew examples. 47 For an analysis of why this shift took place, see section 3.1 below.

462

Rebecca Hasselbach-Andee

tracts. Most of the examples reflect the same formulaic expression ‘I (hereby) bless you by DN’:48 brk-t-k

l-ptḥ by-DN

bless.PERF-1CS-2MS ‘I bless you by Ptaḥ’ (HermB 3:1) brk-t-ky bless.PERF-1CS-2FS

l-ptḥ

by-DN ‘I bless you by Ptaḥ’ (HermB 1:2; 2:2; 4:2; 6:1; 8:1-2) The same formulaic expression is attested in Imperial Aramaic where it occurs in the greeting formula in letters, as in:49

brk-t-ky l-ptḥ by-DN bless.PERF-1CS-2FS ‘I hereby bless you by Ptaḥ’ (A2 1:2) brk-t-k l-yhh w-l-ḥnwm by-DN and-by-DN bless.PERF-1CS-2MS ‘I (hereby) bless you by Yahu and Chnum’ (TAD D7.21:3) hwmy-t-k bl w-nbw and-DN adjure.PERF-1CS-2MS DN ‘I (hereby) adjure you by Bel and Nabu’ (KAI 318:2b-3) Imperial Aramaic further has examples of explicit performatives in contracts:50

48 Hug 1993: 116. As in many other grammatical treatments, Hug cites instances of epistolary perfects together with explicit performatives. 49 Muraoka and Porten 2003: 194; Gzella 2004: 213. See also Kutscher 1971: 110-111. 50 For explicit performatives in Imperial Aramaic see specifically Gzella 2004: 207213.

Explicit performative utterances in Semitic

w-t-ʾmr

šnʾ-t l-ʾsḥwr dismiss.PERF-1CS ACC-PN

463

bʿl-y husband-1CS

and-3FS-say.IMPF ‘should she say: I (hereby) dismiss Esḥor, my husband’ (TAD B2.6-22-23) So far, all the cited examples of explicit performatives in Aramaic use the perfect. In Biblical Aramaic, however, this type of speech act is expressed by the participle: ʾaḇāhāt-ī mәhôḏēʾ u-mәšabbaḥ ʾelāh ACC-2MS God.CSTR forefathers-1CS thank.PTC and-praise.PTC ‘I thank and praise you, o God of my forefathers’ (Dan 2:23a) l-āḵ

ʾanā I

ʾanā I

nәbûḵaḏneṣṣar mәšabbaḥ û-mәrômēm û-mәhaddar PN praise.PTC and-exalt.PTC and-honor.PTC lә-meleḵ šәmayyāʾ ACC-king.CSTR heaven ‘now I, Nebukadnezzar, praise and exalt and honor the king of heaven’ (Dan 4:34)

kәʿan now

Qumran Aramaic exhibits a mixed situation in which both the perfect and participle are used for explicit performatives, although the use of the participle seems to be the more common construction.51 The only clear example for the use of the perfect Rogland cites in his study on Qumran Aramaic (Rogland 1999) is: w-kʿn qbl-t-k mr-y ʿl prʿw Pharaoh and-now lodge_complaint.PERF-1CS-2MS Lord-1cs against ṣʿn mlk mṣrym PN king.CSTR Egypt ‘and now I lodge a complaint before you, my Lord, against Pharaoh Zoan, king of Egypt’ (1QGenAp 20:13-14) Examples of the use of the participle include:

51 Rogland 1999: 278-280.

464

[hʾ]

Rebecca Hasselbach-Andee

ʾnh mwʿd

b-ʿlyʾ b-mrh by-most.High by-lord

rbwtʾ great

b-mlk by-king.CSTR

[behold] I adjure.PTC kwl ʿ[lmym…] ages all.CSTR ‘[behold,] I adjure (you) by the Most High, by the great Lord, by the king of all a[ges…] (1QGenAp 2:4-5) yʾmyʾ ʾnh l-k b-qdyšʾ rbʾ ACC-2MS by-holy_one great adjure.PTC I ‘I adjure you by the great holy one’ (1QGenAp 2:14) w-kʿn l-kh ʿmrm br-y son-1CS and-now ACC-2MS PN w-bny-hwn ʾnʾ mpqd and-sons-3MP I command.PTC

ʾnʾ mpq[d…] I command.PTC…

w-[bn]y-kh and-sons-2MS

‘and now I command you, ʿAmram, my son […] and I command your sons and their sons…’ (4QTQahat 2:9-10) Syriac likewise attests to both the perfect and participle, although it seems that in translation literature, the use of either perfect or participle is dependent on the Vorlage. Rogland argues that the use of the participle in the New Testament and apocryphal literature reflects the underlying Greek present, while the use of the perfect for translations from Hebrew represents the mechanical rendering of the Hebrew original. He further states that in Syriac texts that do not represent translation literature, the participle is the more common and thus more original form in Syriac.52 Rogland quotes several cases (unfortunately not in full Syriac) for the use of the participle in non-Biblical Syriac literature. A baptismal liturgy provides several examples:53 ‘I, N, renounce thee (kpr ʾnʾ PTC.SING I), Satan’ ‘Thee do we call upon (qryn-n PTC.PL-1CP), O Lord Almighty’

52 Rogland 2001: 244-245. Standard grammars of Syriac such as Nöldeke 1904 do not cite any examples of explicit performatives. 53 See Rogland 2001: 246.

Explicit performative utterances in Semitic

465

‘I adjure thee (mwmʾ ʾnʾ PTC.SING I), unclean spirit’ According to Rogland, a collection of magical text provides further examples:54 ‘By thy name, O God – the hope of all who are bound and oppressed, we (hereby) bind and anathematize (w-mḥrmyn-n and-PTC.PL-1CP), expel (w-ṭrdyn-n and-PTC.PL-1CP)… the pains and maladies which reside in the head…’ Rogland concludes that the participle is the genuine expression of explicit performatives in Syriac.55 Syriac thus conforms to Biblical Aramaic and not to earlier Aramaic dialects such as Old and Imperial Aramaic with regard to this feature. Like Hebrew, Aramaic attests to a transition in the expression of explicit performatives from perfect to participle, with dialects such as Qumran Aramaic reflecting an intermediate stage – although the change is almost completed in Qumran Aramaic as well. Interestingly, some Neo-Aramaic dialects such as the dialect from Maʿlula use the perfect for explicit performatives:56 appall-aḫ amōna safety 1CS.grant.PERF-2MS ‘I hereby grant you safety’ aqsmit 1CS.conjure.PERF ‘I conjure you’

aʿl-aḫ PREP-2MS

aqstićć-aḫ 1CS.beg.PERF-2MS ‘I beg you’

54 Rogland 2001: 246. 55 Rogland 2001: 249. 56 See Correll 1978: 47.

466

Rebecca Hasselbach-Andee

Whether the use of the perfect reflects a retention or an innovation in these cases is difficult to tell. The form of the perfect is an innovation, but the function could be inherited. 2.4 Explicit performatives in Ugaritic Explicit performatives in Ugaritic are expressed by the perfect:57 ank… prʿ-t […] šm-k mdd i[l]… I… proclaim.PERF-1CS… name-2MS beloved.CSTR DN ‘I (myself)… (hereby) proclaim [Yammu] as your name, beloved of Ilu…’ (1.1:IV:18-20) l ASSEV

rgm-t tell.PERF-1CS

l-k zbl to-2MS lord

bʿl DN

tn-t repeat.PERF-1CS

l-rkb to-rider.CSTR

ʿrpt clouds ‘I (hereby) truly tell you, o lord Baʿlu, I (hereby) repeat to the rider of clouds’ (1.2:IV:7-8) (?) alpm s̀s̀wm rgm-t ʿly ṯh (?) 1000 horses command.PERF-1CS to GN ‘I (hereby) command (you), to lead X-1000 horses to Ṯ.’ (2.33:24-25) The following passage from a letter that is followed by a request might likewise reflect an explicit performative:58 iršt request

arš-t request.PERF-1CS

l-aḫ-y to-brother-1CS

‘a request I request of my brother…’ (KTU 5.9) 2.5 Explicit performatives in Arabic Classical Arabic, like most other classical Semitic languages discussed so far, uses the perfect for explicit performatives, as in:59 57 Pardee and Whiting 1987: 4; Smith 1995: 790; Tropper 2000: 714. For the examples see Tropper (ibid.). 58 Pardee and Whiting 1987: 11.

Explicit performative utterances in Semitic

467

ḥalaf-tu swear.PERF-1CS ‘I (hereby) swear’ ʾanšad-tu-ka llāh-a implore.PERF-1CS-2MS God-ACC ‘I (hereby) implore you by God’ ʾaslam-tu embrace_Islam.PERF-1CS ‘I (hereby) embrace Islam’ fa-qad wallay-tu bn-a-ka miṣr-a and-PAST appoint.PERF-1CS son-ACC-2MS Egypt-ACC ‘I (hereby) appoint your son administrator over Egypt’ It is difficult to get detailed information on how explicit performatives are expressed in modern Arabic dialects. Syrian Arabic, as described by Cowell (1964), uses the simple imperfect for this type of utterance:60 b-ūʿd-ak ḥa-ʾәdros IMPF-1CS.promise-2MS FUT-1CS.study ‘I promise you I am going to study’ b-әnṣaḥ-ak IMPF-1CS.advise-2MS

əә

nsā-ha forget-3FS

‘I advise you (to) forget it’ It seems that at least some modern Arabic dialects underwent a change in the expression of explicit performatives from perfect to imperfect. 2.6 Explicit Performatives in Ethiopian Semitic Classical Ethiopic or Geʿez uses both the perfect and imperfect for explicit performatives:61 59 See Wright 1898 II: 1; Denz 1982: 71; Fischer 1987: 91; Mayer 1976: 188. 60 Cowell 1964: 325.

468

Rebecca Hasselbach-Andee

Perfect for performatives: nāhu rassay-ku ʾamlāk-a la-farʿon Lord-ACC to-Pharaoh behold place.PERF-1CS ‘behold, I hereby make you (like) the Lord over Pharaoh’ (Ex 7:1) ʾәsma kama-zә ʾamar-ku-ka advise.PERF-1CS-2MS because like-this.MS ‘because of this I advise you…’ (2 Kings 17:11) ʾәsma maḥal-ku for swear.PERF-1CS

ba-ʾәgziʾab(ә)ḥer

by-God ‘for I swear by the God’ (2 Kings 19:7) bārak-nā-kәmmu ba-sәm-a ʾәgziʾab(ә)ḥer by-name-CSTR God bless.PERF-1CS-2MP ‘we bless you in the name of God’ (Ps 128:8) nāhu śem-ku-ka yom la-ʾaḥzāb wa-la-nagaśt behold appoint.PERF-1CS-2MS today to-nations and-to-kings ‘I appoint you today over nations and kings’ (Jer 1:10) Imperfect for performatives: ʾә-sәʾʾәl-aka ʾәgziʾ-әya 1CS-beg.IMPF-2MS Lord-1CS ‘I beg you, my Lord’ (Ex 32:31) wa-ʾә-mәḥḥәl ba-yamān-әya and-1CS-swear.IMPF by-right-1CS ‘and I swear by my right’ (Dtn 32:40) nā-sammәʿ la-kәmmu samāy-a wa-mәdr-a to-2MP heaven-ACC and-earth-ACC 1CP-call_upon.IMPF ‘we call upon heaven and earth as witness against you’ (Jdt 7:28)

61 For these and more examples see Weninger 2001: 79-80; 129-130; see also Tropper 2002: 184.

Explicit performative utterances in Semitic

ʾә-bārrәka-ka 1CS-praise.IMPF-2MS

469

ʾәgziʾ-o Lord-VOC

‘I praise you, o Lord’ (ParJer 7:12) In an important article, Weninger (2000) investigates the use of the perfect and imperfect in Geʿez performatives. He states that the use of perfect versus imperfect is not lexically conditioned but that the two forms stand in free variation (see, e.g. the examples containing the verb bāraka cited above) – although the majority of cases uses the perfect.62 As is well known, most Geʿez texts are translations that are, depending on period, either translated from Greek or Arabic.63 In texts that have a Greek Vorlage, the Geʿez translation is, according to Weninger, deeply influenced by the Greek original. The Geʿez perfect presents the normal rendering of the Greek perfect and aorist, while the use of the imperfect for performatives in Geʿez usually renders the Greek present tense.64 It is difficult to assess from this translated material which verbal form is original to explicit performatives in Geʿez. Weninger argues that there are instances in which the Greek present tense is rendered by the Geʿez perfect. No examples, however, are found in which the Greek aorist or perfect is translated by the Geʿez imperfect. Weninger concludes that the translators used the imperfect to render performatives only under the influence of the Greek Vorlage. This means that in Axumite Geʿez, the normative verbal form used for explicit performatives was the perfect.65 The examples and arguments brought forth by Weninger in favor of his interpretation seem well-founded and convincing. The present study thus follows his analysis and assumes that Geʿez, like many other classical Semitic languages cited so far, uses the perfect to express explicit performatives, while the use of the imperfect is influenced by Greek. In Amharic, the situation differs from Geʿez. Amharic has instances of performatives that occur in either the perfect or imperfect. It seems, however, that in this case, the choice of perfect versus imperfect is lexically determined.66

62 Weninger 2000: 94. 63 The earliest genuine Geʿez texts attested in the Axumite inscriptions contain no cases of explicit performatives (Weninger 2000: 95). 64 Weninger 2000: 96-97. 65 Weninger 2000: 98-99; see also Weninger 2001: 81. 66 For a study of Amharic performatives see Manahlot 1988.

470

Rebecca Hasselbach-Andee

Verbs like ‘to order’, ‘to allow’, ‘to award’, ‘to name’, ‘to appoint’, etc. are usually construed with a perfect, as in:67 hid ɨndɨ-tɨ-hed ɨžž-e-h-all-ähu COMP-2MS-go order.PERF-1CS-2MS.ACC-AUX-1CS go.IMP.MS ‘I have ordered you to go, go!’ = ‘I order you to go’ As in the example just given, these constructions can be followed by an imperative in order to mark them as performatives. Verbs such as ‘to beg’, ‘to request’, ‘to remind’, ‘to swear’, ‘to thank’, ‘to pray’, etc., on the other hand, are construed with the imperfect:68 l-ɨ-sätʾ-ɨh kʾal ɨ-gäba-all-ähu COMP-1CS-give.IMPF-2MS.ACC word 1CS-enter.IMPF-AUX-1CS ‘I promise to give you 5 birr’ (lit. ‘I give you my word to give you 5 birr!’) ammɨst bɨrr five birr

Manahlot states that in the first cases the speaker seems to be of higher social status than in instances of the second type. The second type further primarily occurs at the end of formal business letters and only rarely in actual conversation. Manahlot explains the use of the imperfect in Amharic as: “The writer appeals for some kind of decision which has to be made in the future. Hence, the usage of the imperfective form of the verb becomes pertinent.”69 When the perfect is used, according to Manahlot, the decision has to be made in the mind of the speaker before the utterance and thus the perfective is used.70 The interesting feature of Amharic is the occurrence of both perfect and imperfect for explicit performatives, the use of which seems to be lexically determined. 67 Manahlot 1988: 624. Leslau states that with certain verbs the perfect may express the present, “especially if the action occurs at the moment of speaking” (Leslau 1995: 290). However, he does not cite examples of performatives under the use of the perfect as far as I can tell. 68 Manahlot 1988: 625-626. The form of the imperfect is that of the “compound imperfect” found in Leslau 1995. The basic use of the compound imperfect is for the present and future in main clauses (Leslau 1995: 344). 69 Manahlot 1988: 626. 70 Manahlot 1988: 626.

Explicit performative utterances in Semitic

471

2.7 Explicit performatives in Mehri Mehri uses the perfect for explicit performatives:71 bes,

fsâḫ-en

be-śáraḥ! PREP-conversation

enough stop.PERF-1CP ‘enough, we hereby stop the conversation!’ xalya-k tēš divorce.PERF-1CS ACC.2FS ‘I divorce you (fs)’ hāras-k bayš PREP.2FS marry.PERF-1CS ‘I marry you (fs)’ bār

gzim-k l-ūk t-wōgah hibεr PART swear.PERF-1CS to-2MS 2MS-take.out.IMPF camels ‘I swear you (ms) [must] take the camels out in the early morning’ 2.8 Summary of the Semitic evidence The evidence for explicit performatives in Semitic can be summarized as follows:

Akkadian

Preterite x

Perfect x

Imperfect Participle – –

Hebrew Aramaic

– –

x x

– –

x x

Ugaritic Arabic

– –

x x

– x

– –

Ethiopian Mehri

– –

x x

x –

– –

In Akkadian, the perfect does not reflect the same verbal form as the West Semitic perfect (qatala) but constitutes an inner-Akkadian innovation (iptaras). As mentioned in section 2.1 above, the perfect in Akkadian took over many 71 Wagner 1953: 41; Watson 2012: 96.

472

Rebecca Hasselbach-Andee

functions of the original past/perfective form, the preterite (iprus). A form equivalent to the Akkadian preterite is only attested in vestiges in West Semitic, e.g. in the waw-consecutive of Biblical Hebrew and in the construction lam yaqtul that negates the past in Classical Arabic. It is thus no surprise that the preterite is not attested for performatives in West Semitic, where the innovative perfect qatala took over the functions of the original preterite. Already a cursory glance at the table above indicates that the West Semitic perfect is used for explicit performatives in all WS languages of our sample. No other verbal form exhibits a similarly widespread distribution. Two languages/ language groups also attest to the use of the imperfect (Arabic, Ethiopian), while two others have attestations for the participle (Hebrew, Aramaic). In these four languages/language groups, the use of a non-perfect form for explicit performatives only occurs in later stages of the languages in question while earlier stages use of the perfect. The difference in the expression of explicit performatives in these cases is thus a diachronic one.72 The same is true for the change from preterite to perfect in Akkadian. Based on the widespread distribution of the perfect for explicit performatives in West Semitic languages and the clearly diachronically motivated variations in some sub-groups of West Semitic, it is possible to conclude that the perfect reflects the main and most original means to express explicit performatives in West Semitic. In Akkadian, this role was expressed by the preterite, which was subsequently substituted by the perfect. In both West and East Semitic, the main way to express explicit performatives was thus a past tense/perfective verbal form. A reason why some of the quoted languages underwent changes will be provided in section 3.1 below. 3 Previous interpretations of explicit performatives in Semitic As mentioned in the introduction, since performatives are defined as performing an act by uttering a sentence, they naturally coincide with the present moment and are therefore often described as speech acts that are prototypical for the present and, consequently, for being expressed by present tense forms.73 In many languages, as can be exemplified by English and German sentences of the type “I hereby baptize you…”, “hiermit ernenne ich Sie…”, explicit performatives are in fact expressed by a verbal form used for present tense contexts. 72 For the same assumption see also Gzella 2004: 214. 73 See, e.g., Comrie 1985: 37; Smith 1997: 111.

Explicit performative utterances in Semitic

473

Already Koschmieder noted, however, that some languages use other than present-tense forms for performatives. Greek (both classical and modern), for example, primarily uses the Aorist for this type of utterance. Koschmieder gives the example of a person who is asked “do you admit that you have done X?”, to which the person answers ʾανεγνώρισα ‘I admit (Aorist) it!’.74 Turkish, depending on context, uses either the present, the Aorist, which is used for actions outside a specific time reference, or the II present for performatives, although the Aorist seems to be the most frequently attested form. For performatives, the Aorist is attested in sentences such as rica ederim ‘I beg’ and tebrik ederim ‘I congratulate’.75 He further cites Hebrew as a language that does not use a presenttense form. The use of the Aorist for performatives in Greek has been explained by the fact that this verbal form is used for punctual actions.76 Koschmieder, however, does not agree with this explanation. He argues that punctuality cannot be the reason for using the Aorist in Greek since languages that do not use the present tense for performatives do not necessarily use a verbal form that expresses punctual actions. In fact, as Koschmieder argues, the forms used in Semitic, Turkish, and Slavic do not seem to overlap on a functional level.77 Even languages that use a present tense form do not seem to have any connection with punctuality. In German, for example, performatives can only occur when the particle ‘hiermit’ can be inserted into the sentence, but not temporal adverbs such as ‘jetzt, eben, gerade’, which mark punctual actions of the present. Furthermore, performatives only occur with verbs of speaking (in the broadest sense), have to be in the first person, and are limited to the present tense. According to Koschmieder, these limitations contradict an analysis of performatives as reflecting punctual actions.78 The analysis of the performative as reflecting punctual action is thus incorrect. As an alternative explanation, Koschmieder, based on the theory of K. Bühler (1918), suggests that the “Koinzidenzfall” or performative is part of what Bühler calls “Auslösung”, that is, the resolution or completion of an action designated by constructions such as imperatives,

74 Koschmieder 1945: 23. 75 Koschmieder 1945: 27. 76 See the description in Koschmieder 1945: 23. 77 Koschmieder 1945: 27. 78 Koschmieder 1945: 23.

474

Rebecca Hasselbach-Andee

wishes, requests, etc.79 Koschmieder goes even further in claiming that performatives are the “Auslösung” or resolution/completion of the action itself.80 Koschmieder’s rejection of performatives as punctual acts and his alternative explanation of these utterances as representing “Auslösung”, which has often been equated with completion, has been adopted widely within Semitic studies.81 For example, Weninger (2001) assumes that the perfect is used in many Semitic languages because the action is completed when it is uttered and thus a verbal form is used that expresses the perception of the completed past.82 The idea that performatives use a past tense/perfective verbal form in Semitic languages because they reflect a completed action is in fact already prevalent in Semitic studies before Koschmieder’s analysis and represents the most frequently found explanation in reference grammars of individual Semitic languages. It is closely connected to the notion that the perfect stands for completed actions of the past that is common in early and even present studies of individual Semitic languages.83 A much less commonly found explanation for the use of the preterite/perfect for performatives goes back to the notion of punctuality rejected by Koschmieder. A few reference grammars/grammatical studies state that performatives reflect instantaneous or punctual actions that were expressed by the perfect since the use of the imperfect or participle is not possible for punctual actions.84 The most common interpretation of the Semitic data is thus that performatives are expressed by the preterite/perfect because they reflect completed actions, which can only be expressed by these verbal forms in most Semitic lan-

79 Bühler distinguishes three main functions of language: 1. “Kundgabe”, which designates everything conveyed by means of voice, articulation, choice of words, that is, the utterance itself; 2. “Appell/Auslösung”, meaning the utterance seeks resolution or completion, as in the case of imperatives, wishes, requests; 3. “Darstellung”, meaning the description of situations (Koschmieder 1945: 27). 80 Koschmieder 1945: 28. 81 See Wagner 1997: 58-61 for the reception of Koschmieder’s theory in Hebrew and Semitic studies. Wagner cites, among others, Heimpel and Guidi 1969, Denz 1971, and Mayer 1976 as having been influenced by Koschmieder. 82 Weninger 2001: 78. The same explanation is found in Gzella’s study of Imperial Aramaic, in which he treats performatives with some detail (Gzella 2004: 214). 83 See, e.g., Wright 1898 Vol.2: 1; Dillmann and Bezold 1907: 168; Gesenius and Kautzsch 1910: 311-312; Fischer 1987: 91; Denz 1982: 71. 84 See, e.g., Joüon and Muraoka 1996: 362 and Waltke and O’Connor 1990: 488.

Explicit performative utterances in Semitic

475

guages. This interpretation is closely connected to an aspectual understanding of the respective Semitic verbal systems, with completion being regarded as one of the main defining features of perfective aspect. As mentioned in the introduction, the notion of completion with regard to performatives faces two main problems: 1. the fact that the action is performed while pronouncing the utterance, marking it as prototypical present and complicating the analysis of performatives as “completed” actions, and 2. the outdated notion of perfective aspect as being defined by “completion”. These two issues will be addressed in the following section. 3.1 An alternative analysis Because of the aforementioned issues with an understanding of explicit performatives as reflecting “completed actions”, it is worthwhile to look for an alternative explanation for the use of verbal forms in Semitic. Theoretically, there are two main possibilities: Semitic languages either use a past-tense/perfective verbal form because of the tense value these forms carry (similar to the epistolary perfect) or because of their aspectual value. “Tense”, as is well known, is the grammaticalized expression of location in time, with the most common cross-linguistically grammaticalized notions being simple anteriority, simultaneity, and posteriority.85 In an absolute tense system, which uses the present moment as deictic center, this means that there are three absolute tenses: a past, present, and future.86 Of these three, only the categories of past and present are relevant for the topic of our investigation. It is relatively rare for a situation to truly coincide with the present moment. Most situations either begin before and/or extend longer than the present moment. Since performatives are some of the few utterances that truly coincide with the present moment, from a purely tense-based perspective, we would expect them to be expressed by a verbal form associated with present tense, such as the imperfect or participle. This is the case in some modern Semitic languages, but, as mentioned in section 2.8 above, these instances do not represent the original or most frequently attested situation in Semitic.

85 Comrie 1985: 9, 11. 86 Comrie 1985: 36. The deictic center, that is, whether or not it is the present moment, is not relevant for the present discussion since the preterite/perfect would still reflect anteriority in a relative tense system.

476

Rebecca Hasselbach-Andee

“Past tense” or anteriority locates a situation prior to the present moment. It does not, however, say anything about whether the past situation is a single point in time or an extended period. It equally does not indicate by itself if the anterior situation extends all the way up to the present moment or might even continue into the present or future.87 This, admittedly basic, definition of past tense shows the inherent problem with a tense-based interpretation of the use of the preterite/perfect for explicit performatives: performatives by their very definition do not exhibit any characteristics that would connect them to the notion of anteriority. Nor does a basic understanding of “past tense” provide any explanatory framework for notions such as completion or punctuality. The use of the preterite/perfect in Semitic for explicit performatives consequently cannot be based on the tense value of these verbal forms and has to lie in their aspectual function.88 As mentioned in section 3 above, “aspect” in Semitic has traditionally been associated with the concept of completion – with perfective aspect expressing “completed” and imperfective aspect “incomplete” actions.89 The notion of aspect as being defined by the state of completion of an event is, of course, an outdated understanding of the concept. A more current definition of the term aspect is, in Comrie’s words, to view “the internal temporal constituency of a situation”.90 “Perfective” aspect is defined as representing the 87 Comrie 1985: 41. 88 Assuming a connection to aspect does not mean that the verbal system of the respective Semitic languages was purely or even primarily aspectual. Aspect and tense often combine in the verbal system of a language (see, e.g., Comrie 1976: 9; Smith 1997: 97). In my opinion, such a type of connection is also what underlies most Semitic languages. In other words, the verbal forms of Akkadian, Classical Arabic, Geʿez, and probably an early form of Hebrew and Aramaic, expressed both tense and aspect. 89 See, e.g., Fischer 1987: 91 for the notion that the Classical Arabic perfect designates completed action (= perfective aspect); Dillmann and Bezold (1907: 166), who state that Geʿez has a twofold division of time in which every event is presented as either “finished” or “unfinished”; Nöldeke (1904: 202), who states that the perfect in Syriac denotes past action and completed results; Gesenius and Kautzsch (1910: 125) who describe the Hebrew/Semitic perfect as denoting “that which is concluded, completed, and past” while the imperfect denotes that which is unfinished and/ or continuing. Gesenius and Kautzsch go so far as to claim that Semitic has no tense system but a system that views occurrences from the perspective of completed or incomplete actions (ibid.). 90 Comrie 1976: 3.

Explicit performative utterances in Semitic

477

totality of a situation with no reference to its internal temporal constituency, that is, it represent a situation as unanalysable whole.91 “Imperfective” aspect, on the other hand, makes explicit reference to the internal temporal constituency of a situation with no reference to its beginning or end.92 Comrie stresses that viewing a situation as a whole, that is, as a “complete” action cannot be equated with considering it a “completed” action. Viewing it as a whole can include the beginning and end of a situation, but it does not do so necessarily, as in the Greek use of some perfectives, in which the perfective can express ingressive actions, such as “I became king”, where it expresses the beginning of the situation.93 Comrie further points out that perfective forms are often perceived as indicating situations of short duration and imperfective forms those of long duration, seemingly connecting the two aspects to punctuality and durativity respectively. However, in a sentence such as “He ruled for thirty years”, “ruled” is seen as a whole and is thus representing perfective aspect, although the event itself clearly lasted over a long period of time.94 The notion that the perfective indicates punctual or momentary situations is thus incorrect.95 The impression that perfective aspect describes punctual actions comes from the fact that it views situations as a whole. This can have the effect of reducing them to a single point but does not mean that the action/situation itself is punctual. Although Comrie’s work on aspect has been very influential, it clearly has its limitations in its ability to integrate certain notions such as completion or punctuality from the perspective of either tense or aspect. Throughout the years after the publication of Comrie’s book, the understanding of aspect has been refined and been modified to include such concepts. Smith (1997), for example, proposes that aspect consists of two components. The first component is what she calls “viewpoint”, which corresponds to Comrie’s basic definition of perfective and imperfective aspect. Smith distin91 Comrie 1976: 3. 92 Comrie 1976: 4. An often quoted sentence to show the two aspectual notions of perfective and imperfective in English is “John was reading when I entered”. In this sentence, “was reading” is imperfective since there is no indication of when John started or stopped to read and the action of reading is co-occurring with “entered”. The phrase “he entered”, on the other hand, reflects perfective aspect since it expresses the totality of the event and views it from the outside. 93 Comrie 1976: 18. 94 Comrie 1976: 16. 95 Comrie 1976: 17.

478

Rebecca Hasselbach-Andee

guishes three “viewpoints”: perfective, imperfective, and neutral.96 The second component, which she calls “situation aspect” or “situation type”, classifies the event or state talked about with regard to its temporal properties. In the sentences (1) “Mary walked to school”, (2) “Mary was walking to school”, and (3) “Mary walked in the park”, the first sentence presents a completed event with a clearly stated goal that is reached. The second sentence presents a part of the same event, while the third sentence presents a complete event but the event has no goal.97 It is these types of temporal properties that Comrie’s definition of aspect cannot distinguish while Smith’s two-component approach tries to integrate such notions in a theory of aspect. Smith distinguishes five “situation types”:98 1. State, which is static and durative (‘know the answer’, ‘love Mary’) 2. Activity, which is dynamic, durative, and atelic (‘to laugh’, ‘stroll in the park’) 3. Accomplishment, defined as dynamic, durative, telic, consisting of process and outcome (‘build a house’, ‘walk to school’, ‘learn Greek’) 4. Semelfactive, which is dynamic, atelic, and instantaneous (‘to knock, ‘to tap’) 5. Achievement, which is dynamic, telic, and instantaneous (‘win a race’, ‘reach the top’) “Telic” versus “atelic” distinguishes whether (telic) or not (atelic) the event has a natural endpoint, a goal, outcome, or other change of state. The other two main features that are distinguished in this approach are whether a situation is “static” or “dynamic”, and whether it is “durative” or “instantaneous”. The three main properties that distinguish situation types are thus dynamism, telicity, and duration. These basic distinctions mean that activities, for example, can terminate or stop, but they do not “finish” since they do not have a natural endpoint or goal. Accomplishments consist of single-stage events with no result or outcome. Semelfactives are, as Smith phrases it, the simplest type of events in that they are intrinsically bounded. They usually occur quickly and have no result other than the occurrence of the event. Achievements are instantaneous events that result in a change of state and that are true only for the moment of the event.99 96 Smith 1997: 3. 97 Smith 1997: 2. 98 Smith 1997: 3. 99 Smith 1997: 23, 26, 29, 30.

Explicit performative utterances in Semitic

479

According to the characterization provided by Smith, semelfactives and achievements are instantaneous and have no internal structure. These two categories are thus closely connected to perfective viewpoint, although it has to be stressed that situation type is not necessarily tied to a particular viewpoint.100 When a sentence is formed, the speaker must choose one of each of the two components, that is, a viewpoint and situation type. Viewpoint aspect is commonly signaled by grammatical morphemes that contrast, often in the verbal inflection, while situation aspect is expressed by verb constellations. This means that the latter is not commonly marked overtly in the morphology.101 The two-component analysis of aspect presented by Smith is better suited to provide an explanatory framework for certain situational categories that apply to explicit performatives than Comrie’s definition of aspect. This article thus follows Smith’s extended framework. When looking at explicit performative utterances, we have to decide which aspectual viewpoint and situation type they represent. Explicit performatives reflect a situation/action as a whole. In an utterance such as “I (hereby) promise you I will come tomorrow”, the promise is viewed in its entirety with no reference to its internal temporal constituency. This fact connects them more closely to perfective than imperfective viewpoint. With regard to situation type, explicit performatives, as stated before, are viewed as prototypical present situations. This means that they occur in the moment of speaking and can thus be understood as punctual/instantaneous since they do not extend beyond the moment of speaking. Important in the context of the present investigation is Comrie’s claim that punctual actions, which he defines as a situation that does not last in time but that takes place momentarily – corresponding to Smith’s “instantaneous” category –, has, by definition, no

100 Smith 1997: 72. Smith states that perfective viewpoints tend to present a situation as punctual because of the closed nature of the perfective presentation, although she likewise states that not all perfectives are punctual (ibid.). This analysis contradicts Comrie’s interpretation of perfective aspect. In this point, I tend to agree with Comrie. Although perfective aspect presents a situation as a whole, and thus can give the impression of a punctual situation, this rather seems to be a secondary effect of the primary function of the perfective than an actual function of the perfective that can be generalized. 101 Smith 1997: 4.

480

Rebecca Hasselbach-Andee

internal structure. In languages that have an imperfective form, punctuality and imperfective are thus incompatible.102 Furthermore, explicit performatives are dynamic (i.e. they do not represent states), and they clearly have a natural goal (a promise, appointment, etc.), and are thus telic. In terms of situation type as defined by Smith, explicit performatives thus represent achievements. In my opinion, the fact that explicit performatives can be characterized as achievements, and thus as dynamic, telic, and instantaneous, might be the primary reason why they are expressed by the preterite/perfect in most Semitic languages. Given the fact that they are instantaneous or punctual means that they do not have an internal structure, which, as mentioned above, makes them incompatible with imperfective aspect. They are also telic. This means that there is a sense of “completion”, although a telic situation by itself does not require perfective aspect (as in “He was winning the race”), nor does a dynamic situation (as in “Emily was pushing the cart”).103 The major feature that connects achievements to perfective aspect is thus, as stated by Smith and, in other words, by Comrie, the notion of punctuality/instantaneousness.104 Consequently, the determining factor for the expression of explicit performatives by “perfective” verbal forms in Semitic languages is their instantaneous/punctual character, not the fact that they are telic. This article thus proposes that explicit performatives are expressed by the preterite/perfect in Akkadian and by the perfect in many West Semitic languages because they represent instantaneous/punctual acts. The use of these verbal forms for explicit performatives has nothing to do with any notion of completion as claimed by Koschmieder and others. A number of Semitic languages, however, experienced a shift in verbal forms expressing explicit performatives throughout their history. These lan-

102 Comrie 1976: 42. Other features that have been cited as making performatives incompatible with imperfective aspect are the fact that they constitute individual actions, that is, they cannot be iterative or durative, and that they cannot be modal (Weninger 2001: 77). 103 Smith 1997: 44-46. See also Smith 1997: 234 for the use of the imperfective in Russian for “closed” events such as the sentence zimnij dvorec stroil Rastrelli ‘Rastrelli built (IMPF) the Winter Palace’. 104 Comrie 1976: 42; Smith 1997: 72.

Explicit performative utterances in Semitic

481

guages include Akkadian, Hebrew, Aramaic, Arabic, and Amharic.105 In Akkadian, the perfect took over the functions of the preterite in main clauses (see section 2.1 above). The change from preterite to perfect thus reflects a morphological, not a semantic or functional change. In Hebrew and Aramaic, the participle took over the previous function of the perfect for expressing explicit performatives, while in some modern Arabic dialects and in parts in Amharic, the imperfect replaced the perfect in this specific function. In these languages, the change in verbal forms does reflect a functional change. While the forms themselves are not problematic – they reflect the normative forms used for present tense contexts in the respective languages/stages of the languages and thus conform to what we would expect in terms of expressing explicit performatives – the motivation for these shifts still requires further explanation. It is the thesis of this paper that the shift in verbal forms was caused by a shift in the tense-aspect system of the respective languages. As mentioned above in this section, the reason for the use of the preterite/ perfect for explicit performatives in Semitic languages is assumed to be their instantaneous character, which connects them to an aspectual function of the respective verbal forms. It is well-known that Hebrew underwent a significant change in its verbal system from Biblical to Mishnaic Hebrew (and ultimately modern Hebrew), in which the participle increasingly acquired present-tense function and the perfect and imperfect were reduced to indicating tense, the perfect past-tense, and the imperfect future tense.106 A similar development occurred in Aramaic, where it is noticeable in dialects such as Syriac.107 Syrian Arabic likewise reflects a different verbal system than Classical Arabic. Syrian Arabic developed a new verbal form with prefixed b-, which is used like the English simple present and for the future, that is, like the participle in Hebrew and Aramaic, this form took over the function of present-tense. It is also the form underlying explicit performatives.108 The perfect in Syrian Arabic indicates

105 Geʿez is not considered in this context since the use of the imperfect for explicit performatives seems to be motivated by the underlying verbal form of the original text in translation literature. 106 See, e.g., the description of the use of the major tenses in modern Hebrew in Coffin and Bolozky 2005: 35-38. 107 Nöldeke 1904: 202, 208, 211. 108 Cowell 1964: 325. Syrian Arabic developed another verbal form based on the prefix conjugation, which exhibits the prefix ʿam-. This form designates the state or

482

Rebecca Hasselbach-Andee

that an event is in the past. The verbal system of Syrian Arabic is thus tensebased without any significant notion of aspect (in its main verbal forms). The languages that underwent a shift in the expression of explicit performatives consequently also underwent a shift in the function of the respective verbal forms, in which they became primarily tense-based and lost most or all of their aspectual connotations. The loss of the respective aspectual function made the use of the perfect(ive) for instantaneous actions inadequate and these languages shifted to the verbal forms used for present tense, that is, the participle or simple imperfect respectively. 4

Conclusions

The present article has argued that the common understanding that explicit performatives are expressed by the preterite/perfect in classical Semitic languages because they express completed actions should be revised because completion or telicity are not necessarily features associated with perfective aspect. Instead, these utterances are better understood as instantaneous/punctual. The use of the preterite/perfect reflects the close connection of perfective aspect with this temporal property. This also means that the verbal system of the respective languages must, at least in part, reflect aspect in addition to tense. The shift of verbal forms used for explicit performatives in some Semitic languages, such as Hebrew, Aramaic, and Arabic, can be explained by changes in the tense-aspect system of the respective languages. In these cases, aspectual function was lost, or at least weakened, and the languages shifted to expressing explicit performa-tives by a form marking present-tense. It is important to note that all the lan-guages in question underwent the same underlying shift (perfective > present tense), although its representation can differ depending on which form became the main means to express the present (participle versus imperfect). Consequently, the expression of explicit performatives in Semitic languages is closely connected to the tense-aspect system of the respective languages and sensitive to changes in this system.

activity going on at the moment, like the English progressive form (Cowell 1964: 320), and thus cannot be used for explicit performatives, which are instantaneous.

Explicit performative utterances in Semitic

483

Bibliography Austin, John L. 1962. How to do Things with Words. Oxford: Clarendon. Coffin, Edna A. and Shmuel Bolozky. 2005. A Reference Grammar of Modern Hebrew. Cambridge: Cambridge University Press. Comrie, Bernard. 1976. Aspect: An Introduction to the Study of Verbal Aspect and Related Problems. Cambridge: Cambridge University Press. Comrie, Bernard. 1985. Tense. Cambridge: Cambridge University Press. Correll, Christoph. 1978. Untersuchungen zur Syntax der neuwestaramäischen Dialekte des Antilibanon. Wiesbaden: Franz Steiner. Cowell, Mark W. 1964. A Reference Grammar of Syrian Arabic. Washington D.C.: Georgetwon University Press. Crystal, David. 2003. A Dictionary of Linguistics and Phonetics (5th edition). Malden: Black-well. Dillmann, August and Carl Bezold. 1907. Ethiopic Grammar (reprint 2003). Eugene: Wipf and Stock. Denz, Adolf. 1982. Die Struktur des Klassischen Arabisch. In Grundriß der Arabischen Philologie – Band 1: Sprachwissenschaft, Wolfdietrich Fischer (ed.), 58-82. Wiesbaden: Ludwig Reichert. Fischer, Wolfdietrich. 1987. Grammatik des Klassischen Arabisch. Wiesbaden: Harrassowitz. Gesenius, Wilhelm and Emil Kautzsch. 1910. Gesenius’ Hebrew Grammar (second English edition). Oxford: Clarendon. Gzella, Holger. 2004. Tempus, Aspekt und Modalität im Reichsaramäischen. Wiesbaden: Harrassowitz. Hasselbach, Rebecca. 2005. Sargonic Akkadian: A Historical and Comparative Study of the Syllabic Texts. Wiesbaden: Harrassowitz. Heimpel, Wolfgang and Gabriella Guidi. 1969. Der Koinzidenzfall im Akkadischen. Zeitschrift der Deutschen Morgenländischen Gesellschaft, Suppl. 1: 148-152. Hillers, Dilbert R. 1995. Some Performative Utterances in the Bible. In Pomegranates and Golden Bells: Studies in Biblical, Jewish, and Near Eastern Ritual, Law, and Literature in Honor of Jacob Milgrom, David P. Wright, David N. Freedman, Avi Hurvitz (eds.), 757-766. Winona Lake: Eisenbrauns. Hug, Volker. 1993. Altaramäische Grammatik der Texte des 7. und 6. Jh.s v. Chr. Heidelberg: Heidelberger Orientverlag. Joüon, Paul and Takamitsu Muraoka. 1996. A Grammar of Biblical Hebrew – Part Three: Syntax. Rome: Pontificio Istituto Biblico. Kienast, Burkhart and Konrad Volk. 1995. Die sumerischen und akkadischen Briefe des III. Jahrtausends aus der Zeit vor der III. Dynastie von Ur. Stuttgart: Franz Steiner. Koschmieder, Erwin. 1945. Zur Bestimmung der Funktionen grammatischer Kategorien. Abhandlungen der Bayerischen Akademie der Wissenschaften 25. München: Verlag der Bayerischen Akademie der Wissenschaften. Kouwenberg, N.J.C. 2010. The Akkadian Verb and its Semitic Background. Winona Lake: Eisenbrauns. Kutscher, Ezekiel Y. 1971. The Hermepolis Papyri. Israel Oriental Studies 1: 103-119.

484

Rebecca Hasselbach-Andee

Leslau, Wolf. 1995. Reference Grammar of Amharic. Wiesbaden: Harrassowitz. Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press. Manahlot, Demissie. 1988. Some Notes on Amharic Performative Verbs. In Proceedings of the Eighth International Conference of Ethiopian Studies, University of Addis Ababa, 1984, (Vol. 1) Taddese Beyene (ed.); 623-628. Addis Ababa: Institute of Ethiopian Studies. Mayer, Werner. 1976. Untersuchungen zur Formensprache der babylonischen “Gebetsbeschwörungen”. Rome: Biblical Institute. Muraoka, Takamitsu and Bezalel Porten. 2003. A Grammar of Egyptian Aramaic (second revised edition). Leiden: Brill. Nöldeke, Theodor. 1904. Compendious Syriac Grammar (reprint 2001). Winona Lake: Eisenbrauns. Pardee, Dennis and Robert M. Whiting, 1987. Aspects of Epistolary Verbal Usage in Ugaritic and Akkadian. Bulletin of the School of Oriental and African Studies 50: 1-31. Recanati, François. 1987. Meaning and Force: The Pragmatics of Performative Utterances. Cambridge: Cambridge University Press. Rogland, Max. 1999. A Note on Performative Utterances in Qumran Aramaic. Revue de Qumran 19: 277-280. Rogland, Max. 2001. Performative Utterances in Classical Syriac. Journal of Semitic Studies 46: 243-250. Smith, Carlota S. 1997. The Parameter of Aspect (second edition). Dordrecht: Kluwer. Smith, Mark S. 1995. The *qatala Form in Ugaritic Narrative Poetry. In Pomegranates and Golden Bells: Studies in Biblical, Jewish, and Near Eastern Ritual, Law, and Literature in Honor of Jacob Milgrom, David P. Wright, David N. Freedman, Avi Hurvitz (eds.), 789-803. Winona Lake: Eisenbrauns. von Soden, Wolfram. 1995. Grundriss der Akkadischen Grammatik (third edition). Rome: Pontificio Istituto Biblico. Streck, Michael P. 1995. Zahl und Zeit: Grammatik der Numeralia und des Verbalsystems im Spätbabylonischen. Groningen: Styx. Tropper, Josef. 2000. Ugaritische Grammatik. Münster: Ugarit. Tropper, Josef. 2002. Altäthiopisch: Grammatik des Geʿez mit Übungstexten und Glossar. Münster: Ugarit. Wagner, Andreas. 1997. Sprechakte und Sprechaktanalyse im Alten Testament: Untersuchungen im biblischen Hebräisch an der Nahtstelle zwischen Handlungsebene und Grammatik. Berlin: de Gruyter. Wagner, Ewald. 1953. Syntax der Mehri-Sprache – Unter Berücksichtigung auch der anderen neusüdarabischen Sprachen. Berlin: Akademie-Verlag. Waltke, Bruce K. and Michael O’Connor. 1990. An Introduction to Biblical Hebrew Syntax. Winona Lake: Eisenbrauns. Watson, Janet C.E. 2012. The Structure of Mehri. Wiesbaden: Harrassowitz. Weninger, Stefan. 2000. On Performatives in Classical Ethiopic. Journal of Semitic Studies 45: 91-101. Weninger, Stefan. 2001. Das Verbalsystem des Altäthiopischen. Wiesbaden: Harrassowitz.

Explicit performative utterances in Semitic

485

Westenholz, Aage and Joan Westenholz. 1977. Help for Rejected Suitors: The Old Akkadian Love Incantation MAD V 8. Orientalia 46:198-219. Wright, Wiliam. 1898. A Grammar of the Arabic Language (reprint 1999) – Vol. 2. Cambridge: Cambridge University Press.

Coordinated clause structures in Scandinavian and Semitic involving a finite verb form and an infinitive Janne Bondi Johannessen, University of Oslo, and Lutz Edzard, University of Erlangen-Nürnberg and University of Oslo

1 Introduction One of the two co-authors, Lutz Edzard, has been wondering for some time about a construction found in Norwegian (and also in other Scandinavian languages), which consists of a finite verb form that is coordinated with an infinitive: (1) Finite verb form coordinated with an infinitive in Norwegian De ble stående og vente wait.INF they become.PRET stand.PRES.PART and ’They remained standing, waiting.’ (Lødrup 2002: 138) The paper consists of three sections. First, we will consider an inner-Scandinavian explanation of the phenomenon in question. Second, we will draw parallels between the Norwegian construction in (1) and similar constructions in Semitic, both in older stages (Phoenician, Biblical Hebrew, and Ancient/Old North Arabian) and in modern Ethio-Semitic. And third, we will explore to what extent these two approaches can be aligned and benefit from each other. 2

An inner-Scandinavian explanation

In order to account for pseudo-coordination, the common term for the construction type to which (1) belongs, we start by giving a brief overview of verbal tenses in Scandinavian, exemplified by Norwegian (2): (2) Verbal tenses in Norwegian a. Finite forms: Present tense: kjører (’drives’) Preterite tense: kjørte (’drove’) Imperative: kjør (’drive’)

Coordinated clause structures involving a finite verb form and an infinitive

487

b. Infinite forms: Infinitive: kjøre (’drive’) Present participle: kjørende (’driving’) Past participle: kjørt (’driven’) (adapted from Faarlund et al. 1997: 468) Aspect is not an inflectional category in Scandinavian; instead this semantic category is expressed periphrastically in a number of ways. Faarlund et al. (1997: §73.3.2) give a brief summary of Norwegian for such aspects as cursive, ’close to’, continuative, habitual, ingressive, egressive, iterative, and resultative. These are all expressed by an extra verb or verb phrase. Pseudo-coordination (in Danish also called the agreement construction1) is one way of expressing durative aspect. Here we will focus on a subgroup of those; those that are expressed with a posture verb. They express durative or progressive aspect. Examples from the three Mainland Scandinavian languages are given in (3) – (5) (in the rest of the paper, we will focus on Norwegian). (3) Pseudo-coordination in Norwegian En mann sitter og leser en bok and read.PRES a book a man sit.PRES ’A man sits reading a book.’ (Lødrup 2014: 43) (4) Pseudo-coordination in Swedish Han sitter o äter he sit.PRES and eat.PRES ’He sits eating.’ (adapted from Wiklund 2007: 115) (5) Pseudo-coordination in Danish Katten lå og sov lie.PRET and sleep.PRET cat.DEF ’The cat lay sleeping.’ (Hansen and Heltoft 2011: 979) The construction seems to contain two equal verbs (here: sitter ’sits’ and leser ’reads’) coordinated by the conjunction og ’and’. Many linguists have the view, though, that this construction has been reanalyzed into a complex predi                                                                                                                 1

 

The corresponding Danish term is ”kongruenskonstruktionen”.

488

Janne Bondi Johannessen and Lutz Edzard

cate, so that it is not to be understood as an ordinary coordination. Faarlund et al. (1997: 534f.) sum up this view by the following arguments: The two activities represented by the two verbs are not equal. The first verb has a bleached meaning, and serves to indicate the duration of the activity of the second verb. The first verb is about to become an auxiliary according to Faarlund et al., in other words a complex predicate. Grammatical tests also reveal that this is no ordinary coordination: The two verbs cannot swap places, as they can in ordinary coordination. If the construction is negated, one negation is enough, after the first verb, as if it were an auxiliary. Two ordinary coordinated verbs would have needed a negation each. And if the sentence is questioned, the subject follows the first verb, not the two verbs, as it would do in ordinary coordination. In some dialects this construction can even leave out the conjunction, something that can never be done in ordinary coordination of two elements. Since pseudo-coordination is somewhat different from coordination proper, many researchers have proposed that there is some kind of reanalysis taking place, in which the structure is reduced. Lødrup (2014: 44-46) sums up the literature on this topic, identifying three main trends: –

The posture verb is not a regular argument-taking main verb, it has some special grammatical status, which is seen in connection with its progressive aspect.

–

The construction is monoclausal, or its subordinate clause is reduced in some way. The posture verbs and the following verb constitute one complex predicate (in some analyses).

–

In Lødrup’s summary, some have suggested that the posture verb is an auxiliary, others that it is a light verb in a reanalysis construction, and some would like to regard the posture verb as the first verb in a serial verb construction. Lødrup, however, does not agree that a reanalysis of the verbs as a complex predicate is the right way of understanding these constructions. First, the posture verb does not lose its literal meaning. It can even be focussed and stressed, as in (6), where it is implied that the subject should be standing: (6) Focussed posture verb in complex predicates Sitter han virkelig og snakker med dronningen? he really and talk.PRES with queen.DEF sit.PRES ‘Is he really sitting when talking to the queen?’ (Lødrup 2014: 47)’

 

Coordinated clause structures involving a finite verb form and an infinitive

489

Another argument given by Lødrup against a one-predicate analysis is the fact that the posture verb and the lexical verb can have different subjects. In (7) the posture verb has a formal subject, while the second verb has a lexical subject controlled by the object of the first verb: (7) Different subjects in complex predicates Det sitter en mann her og leser there

en bok

sit.PRES

a man here and read.PRES a ‘A man is sitting here, reading a book.’ (Lødrup 2014: 48)

book

Lødrup also points out that the two verbs in this construction can have different diathesis. In (8) the first verb is active and the next verb passive: (8) Different diatheses in complex predicates Et barn sitter i stolen og blir in chair.DEF and become.PRES a child sit.PRES ‘A child is fed sitting in the chair.’ (Lødrup 2014: 48)

matet feed.PAST.PART

Further, the two verbs do not need to have the same temporal properties. In (9) the posture verb has present tense, while the second verb has periphrastic future tense. (9) Different tense in complex predicates Nå sitter jeg her og skal levere søknaden I here and shall.PRES submit.INF application.DEF now sit.PRES om fem minutter in five minutes ‘I am sitting here, going to submit my application in five minutes.’ (Lødrup 2014: 51) Lødrup (2014) also gives examples of something that is poorly understood: the obligatory formal agreement between the posture verb and the (first) verb of the second conjunct. Some researchers use this agreement as a final argument for reanalysis, but Lødrup (2014: 59) points out that the agreement requirement in pseudo-coordination is different than other cases of reanalysis. In pseudocoordination it is not only the morpho-syntactic category that has to be the same, but even the actual form. Thus, although there are two ways of forming  

490

Janne Bondi Johannessen and Lutz Edzard

the passive in Norwegian, in pseudo-coordination the same form must be used. Therefore, (10) and (11) are both correct, but choosing one form for the posture verb and another for the second verb is not possible, see (12) and (13). Notice that all the sentences should be understood in the same way: ’One sits producing’ [knitting presents for Christmas]. (10) Possible pseudo-coordination Men det blir sittet og but there become.PRES sit.PAST.PART and

produsert produce.PAST.PART

(11) Possible pseudo-coordination Men det sittes og produseres and produce.PASS but there sit.PASS (12) Impossible pseudo-coordination *Men det sittes og blir and become.PRES but there sit.PASS

produsert produce.PAST.PART

(13) Impossible pseudo-coordination *Men det sittes og produserer and produce.PRES but there sit.PASS Turning now to (1), repeated below, and its Swedish equivalent in (14), this construction seems to violate the formal agreement requirement just shown.2 (1) Finite verb form coordinated with an infinitive in Norwegian De ble stående og vente wait.INF they become.PRET stand.PRES.PART and ’The remained standing, waiting.’                                                                                                                 2

In Danish, this kind of “unbalanced”, non-agreeing pseudo-coordination does not seem to exist. It is not mentioned by Hansen and Heltoft (2011), and google searches only reveal a “balanced”, agreeing type, in which the second conjunct verb has the same inflectional form as the finite verb of the first conjunct. de blev stående og stand.PRES.PART and they become.PRET ‘The remained standing looking at you.’ (cyberhus.dk/brevkasse/seksualitet/busted-)

 

kiggede look.PRET

på jer on you

Coordinated clause structures involving a finite verb form and an infinitive

491

(14) Finite verb form coordinated with an infinitive in Swedish jag blev sittande och titta på bilderna I

become.PRET

sit. PRES.PART and

look.INF

on picture.PL.DEF

’I remained sitting, looking at the pictures.’ (www.annawahlgren.com) In (1) and (14), the posture verb phrase consists of a an auxiliary plus the posture verb, which is a present participle. The first verb after the conjunction is an infinitive. Clearly there is no formal agreement. Lødrup’s (2002: 138) explanation for this, however, is that the posture verb is expressed in something that has been derived from a verb into an adjective. It therefore has no formal agreement features to be inherited by the verb of the second conjunct. One could ask why the latter verb is then infinitive. Lødrup (personal communication) suggests that it is simply because the infinitive is the standard form of the verb, which is used in dictionary entries and therefore natural to choose when there is no form to agree with. Lødrup’s analysis of the pseudo-coordination with posture verbs is a biclausal one, in which the conjunction is a subordinating element (the coordinating conjunction og ’and’ and the infinitival marker å ’to’ have the same form /o/ in all spoken Norwegian), confusingly written as the conjunction in the official norm. The second verb is then the start of a subordinated clause (Lødrup 2014: 45). The reference grammars for Norwegian and Danish (Faarlund et al. 1997 and Hansen and Heltoft 2011, respectively) both suggest a complex predicate analysis, as does Wiklund (2007). However, the arguments for pseudo-coordination and complex predicates do not cover the peculiarities of the construction, especially the tense agreement restrictions. We will therefore instead analyze pseudo-coordination as coordination proper, and show that what is then superficially incompatible with a coordination analysis, i.e. the type exemplified in (1), is not a problem. This analysis also makes it possible to see the similarities between the construction in Scandinavian and some Semitic languages. 3 Typological parallels in Semitic In Semitic, one encounters infinite verb forms in positions where one would expect a finite form from a typical European perspective. In some Semitic languages, infinitives (or verbal nouns) with or without enclitic pronominal suffixes can either function as gerunds, marking subordinate events (or events ”in

 

492

Janne Bondi Johannessen and Lutz Edzard

the background”) or as ”full-fledged” verb forms marking the main event (or event ”in the foreground”) in a sentence. We start by presenting an overview of morphological facts regarding forms with (enclitic) pronominal suffixes. In a historical perspective, the Akkadian (third to first millenium BC) stative is the most obvious case of a conjugated verbal noun. The term “verbal adjective”, as used notably by John Huehnergard, underscores this opacity. Indeed, Huehnergard (2005: 614) subsumes the “Vbl. Adj. + Pron. Subj.” forms under the category “Non-Finite Forms”. Not only verbal adjectives, but also nouns can undergo inflection in Akkadian, e.g., šarrāku ‘I am king’, from šarru(m) ‘king’ (comparable to a regular stative as, e.g., marṣāku ‘I am ill’, from marṣum ‘ill). Here is the paradigm, based on the synchronically underlying form /paris/, belonging to the root √ p-r-s ‘to divide’ (15): (15) Paradigm of the verbal adjective / stative in Akkadian paris 3SG.M parsat 3SG.F parsāta 2SG.M 2SG.F parsāti parsāku 1SG.C parsū 3PL.M parsā 3PL.F 2PL.M parsātunu 2PL.F 1PL.C

parsātina parsānu

Comparable to the the scenario of the Akkadian verbal noun / stative, both nouns and adjectives (including participles) can be followed by clitic pronouns in Classical Syriac (Aramaic), e.g. (cf. Rubin 2005: 31f.) (16):   (16) Clitic pronouns in Classical Syriac wә-ʾena ʿap̄rā (ʾ)nā CONJ-I dust I ‘And I am dust.’ (Gen 18:27) ʾāmar (ʾ)nā lә-ḵōn saying I to-you.PL.M ‘I say to you.’ (Matthew 3:9)

 

Coordinated clause structures involving a finite verb form and an infinitive

493

The original bound orthography in these cases actually supports the analysis of these examples in terms of a cliticization process. The obvious Semitic example of a conjugated nominal form is the EthioSemitic converb. Here is a chart, juxtaposing Gәʿәz (classical Ethiopic) and modern Amharic converb forms, based on the root √ n–g–r ’say’ (cf. also Hetzron 1972: 101) (17). (17) The converb/gerund in Gәʿәz and Amharic Gәʿәz (type A) Amharic (type A) 3SG.M ነጊሮ nägir-o ነግሮ nägro 3SG.F ነጊራ 2SG.M ነጊረከ 2SG.F ነጊረኪ 1SG.C ነጊርየ 3PL.M ነጊሮሙ 3PL.F ነጊሮን 2PL.M ነጊረክሙ 2PL.F ነጊረክን 1PL.C ነጊረነ

nägir-a

ነግራ

nägra

nägirä-kä nägirä-ki nägir-әyyä nägir-omu nägir-on

ነግረህ ነግረሽ ነግሬ ነግረው

nägräh nägräš nägәrre nägräw

nägirä-kәmu nägirä-kәn

ነግራችሁ

nägirä-nä

ነግረን

“ nägraččәhu “ nägrän

In Gәʿәz, there is no doubt that the converb can be analyzed as the infinitivelike noun in the adverbial accusative (or better: the dependent case), followed by possessive suffixes, as clearly evidenced by the forms of the second person; in Amharic, different historical analyses exist, but this is not central to the discussion at hand. We will now consider conjugated infinite forms in context. In Northwest Semitic, the coordination of infinitive-based converbs with finite verb forms occurs not infrequently (cf. also Lipiński 2001: 427) (18): (18) Converb-like constructions in (a) Biblical Hebrew and (b) Phoenician (a) Biblical Hebrew: wa-yhī ka-hărīm-ī qōl-ī wā-ʾeqrāʾ CONJ-be.PRET.3SG.M as-lift.INF-my voice-my CONJ-cry.PRET.1SG ‘Lifting up my voice I cried.’ (Gen 39:18)

 

494

Janne Bondi Johannessen and Lutz Edzard

(b) Phoenician: pʿl ʾnk … l-rbt-y make.INF I … to-lady-my

… w-šmʿ … CONJ-hear.PRF.3SG.F

ql voice(-my)

‘I having made (this) … for my Lady …, she heard my voice.’ In the Phoenician Karatepe Inscription from ca. 700 BC one encounters both perfect forms (Semitic suffix conjugation) and gerund-like forms, which – at least in their prevalent interpretation – consist of absolute infinitive forms followed by enclitic personal pronouns. Here (19) are lines 1–12 of the inscription, in a colometric division (the vocalization is due to Hackett 2013): (19) Karatepe Inscription, Phoenician, (lines 1–12) colon 1: ʾnk ʾztwd hbrk-bʿl ʿbd bʿl ʾanokī ʾazitawadd(a) ha-barūk-baʿl ʿabd baʿl, I am Azitawadda, the blessed one of Baʿl, the servant of Baʿl, colon 2: ʾš ʾdr ʾwrk mlk dnnym ʾaš ʾiddir (make_strong.PRF.3SG.M) ʾawar(i)k(u) milk danūnīyīm. whom Wariku the king of the Danunites made strong. colon 3: pʿln bʿl ldnnym lʾb wlʾm paʿo/al(a)-nī (make.INF-me/PRF.3SG.M-me) baʿl la-danūnīyīm la-ʾab wa-la-ʾimm. Baʿl made me a father and a mother to the Danunites. colon 4: yḥw ʾnk ʾyt dnnym yaḥwi ʾanokī (give_life.INF I)ʾiyyat danūnīyīm, Because I gave life to the Danunites, colon 5: yrḥb ʾnk ʾrṣ ʿmq ʾdn lmmṣʾ šmš wʿd mbʾy yarḥib ʾanokī (broaden.INF I) ʾarṣ ʿimq ʾadana la-mim-moṣāʾ šamš wa-ʿad maboʾiyū, because I broadened the land of the plain of Adana from east to west, colon 6: wkn bymty kl nʿm ldnnym wšbʿ wmnʿm wa-kon/kan (be.INF/PRF.3SG.M) ba-yamotī kull nuʿm la-danūnīyīm, wa-šubʿ wamanʿam. there was in my day all good for the Danunites, and plenty and prosperity, colon 7: wmlʾ ʾnk ʿqrt pʿr wa-malliʾ ʾanokī (fill.INF I) ʿvqrot paʿ(a)r. and I filled the granaries of Paʿr.

 

Coordinated clause structures involving a finite verb form and an infinitive

495

colon 8: wpʿl ʾnk ss ʿl ss wmgn ʿl mgn wmḥnt ʿl mḥnt wa-paʿol ʾanokī (make.INF I) sūs ʿal sūs wa-magin ʿal magin wa-maḥnat ʿal maḥnat And I added (lit.: made) horse to horse, and shield to shield, and camp to camp bʿbr bʿl wʾlm ba-ʿvbūr baʿl wa-ʾilīm. by the grace of Baʿl and the gods. colon 9: wšbrt mlṣm wa-šibbirtī/šabartī (break.PRF.1SG) milīṣīm. And I shattered the villains, colon 10: wtrq ʾnk kl hrʿ wa-taroq ʾanokī (uproot.INF I) kull ha-raʿʿ, and uprooted all the evil colon 11: ʾš kn bʾrṣ ʾaš kan (be.PRF.3SG.M) ba-ʾarṣ. that was in the land. colon 12: wyṭnʾ ʾnk bt ʾdny bnʿm wa-yaṭniʾ ʾanokī (establish.INF I) bet ʾadonī ba-nuʿm. I established the house of my lord in goodness, colon 13: wpʿl ʾnk wšrš ʾdny nʿm wa-paʿol ʾanokī (do.INF I) la-šurš ʾadonī nuʿm. and acted well to the stock (lit.: root) of my lord, colon 14: wyšb ʾnk ʿl ksʾ ʾby wa-yošib-o ʾanokī (place.INF-him I) ʿal kissiʾ ʾabīyū. placing it (him) on its (his) father’s throne. colon 15: wšt ʾnk šlm ʾt kl mlk wa-šot ʾanokī (set.INF I) šalom ʾitt kull milk. And I made peace with every king. colon 16: wʾp bʾbt pʿln kl mlk bṣdqy wa-ʾap ba-ʾabbūt paʿo/al(a)-nī (make.INF-me/PRF.3SG.M-me) kull milk ba-ṣidqī And also, every king regarded (lit: made) me as a father because of my justice, wbḥkmty wbnʿm lby wa-ba-ḥukmatī, wa-ba-nuʿm libbī. my wisdom, and the goodness of my heart.

 

496

Janne Bondi Johannessen and Lutz Edzard

While some cases are ambiguous (cola 3, 6, 16), forms with an absolute infinitive plus pronoun appear to prevail (always in the first person singular). Perfect forms regularly stand in relative clauses (cola 2, 11). The exact use of tense in this document is still under discussion, but the interplay of perfect forms and infinitive-based gerunds is well established. One can easily single out passages that resemble the finite verb form + infinitive scenario under discussion, e.g., colon 9 and 10 (20): (20) Pseudo-coordination in Phoenician colon 9: wšbrt mlṣm wa-šibbirtī/šabartī milīṣīm CONJ-break.PRF.1SG villain.PL ’And I shattered the villains colon 10: wtrq ʾnk kl hrʿ wa-taroq ʾanokī kull CONJ-uproot.INF I all and uprooted all the evil.’

ha-raʿʿ DEF-evil

Especially interesting – in its resemblance to the Norwegian scenario – is the situation in Safaitic (a branch of Ancient North Arabian, attested approximately from the first century BC to the 4th century AD), where the first event in a chain of events may be expressed by a perfect form and the second by a bare infinitive, without any pronominal suffix (cf. Al-Jallad 2015, chapter 16.1) (21): (21) Pseudo-coordination (perfect + infinitive) in Safaitic rʿy h-rmḫ bql w kmʾt DEF-camel.COL herbage CONJ gather_truffles.INF pasture.PRF.3.SG.M ‘He pastured the camels on spring herbage and gathered truffles.’ (MAHB 5) w

wrd f nyt (b-)ʾmtn CONJ go_to_water.PRF.3.SG.M CONJ migrate.INF (in-)Libra ‘And he went to the water, and then migrated when the sun was in Libra.’ (KRS 1770) Brockelmann (1913: 168) argues that similar cases are already found in Biblical Hebrew, e.g., in 1 Sam 2:27–28 (cf. also Rubinstein 1952, Huesman 1956, Waltke & O’Connor 1990: 595ff., among other references) (22):

 

Coordinated clause structures involving a finite verb form and an infinitive

497

(22) Pseudo-coordination (perfect + infinitive) in Biblical Hebrew hă-niḡlō niḡlēṯī ʾel bēṯ ʾāḇī-ḵā INT-be_revealed.INF.ABS reveal.REFL.PRF.1SG.M to house.CS father.CS-you bi-hyōṯ-ām bә-miṣrayim in-be.INF.CS-their in-Egypt

lә-ḇēṯ for-house.CS

ū-ḇāḥōr

mik-kol

ʾōṯ-ō

parʿō(h) Pharaoh

šiḇṭē

yiśrāʾēl …

CONJ-choose.INF.ABS

OBJ-him from-all tribe.PL.CS Israel … ‘Did I not clearly reveal myself to your ancestor's family when they were in Egypt under Pharaoh? And did I not choose him from all the tribes of Israel ...’ (1 Sam 2:27–28)

kī

ṣamtem wә-sāp̄ōḏ fast.PRF.2PL.M CONJ-mourn.INF.ABS ’When you fastened and mourned.’ (Zech 7:5) CONJ

zәraʿtem harbē wә-hāḇē mәʿāṭ sow.PRF.2PL.M much CONJ-harvest.INF.ABS little ‘You have sown much but harvested little.’ (Hag 1:6) Morrison (2013: 266f.) summarizes the relevant discussion regarding Biblical Hebrew with the following arguments: (i) “The infinitive absolute […] can stand for participles and finite verbs while not expressing any particular nuance that would distinguish it from the form that it substitutes.” (ii) “[L]ater scribes may have appealed to the infinitive absolute because they were uncertain about the finite form.” (iii) “The use of the of the infinite absolute to substitute for a finite verb may have developed from the infinitive’s adverbial use that gradually allowed it to stand for itself.” Turning to modern Semitic, comparable examples are wide-spread, notably in Ethio-Semitic Amharic. Here, in a sequence of events, the finite verb form is typically reserved to the last action, whereas previous actions are typically expressed by gerund/converb forms. The converb forms then have to be understood and translated according to the tense and mood of the final verb (cf. Appleyard 1995 and Edzard 2014) (23):

 

498

Janne Bondi Johannessen and Lutz Edzard

(23) Sequence of events in Amharic taksi ṭärtäw täsaffәräw kä-ṭәqit gize taxi call.CVB.3PL get_in.CVB.3PL of-little time

bä-hwala after

mәgәb bet food house

yәdärsallu arrive.IPF.3PL ‘They call a taxi, get in, and after a while they arrive at the restaurant.’ (“having called, a taxi, having gotten in, …”) gäbto täqämmäṭä sit_down.PRF.3SG.M come_in.CVB.3SG.M ’He came in and sat down.’ (“[he] coming in he sat down”) qämis märṭäš gәži-ll-at shirt choose.CVB.2SG.F buy.IMP.SG.F-for-her ‘Choose (sg.f.) a shirt and buy (sg.f.) it for her!’ (“[you (sg.f.)] choosing a shirt buy it for her!”) Without claiming that the verbal syntax of these Semitic examples can be directly compared with the verbal syntax in the Norwegian example under investigation, the typological parallels are nevertheless striking. The ideas about the ”deranking” of (subordinate) finite verb forms expressed by Yri (2014) appear to be pertinent in both the Scandinavian and the Semitic scenario. 4

A syntactic analysis

One must ask how it is possible that languages that are so different from each other can share a common feature that seems so idiosyncratic. Of course, with subordination we expect that the matrix verb and the subsequent verb can have different features of finiteness, and even other morpho-syntactic features like aspect, agreement etc. So if pseudo-coordination in Scandinavian and the sequence of events construction in Amharic are analyzed as subordination or as a complex predicate, there is no problem. This is what Lødrup (2002, 2014), Wiklund (2007) and other have argued for Scandinavian. The Semitic constructions, however, should not be automatically analyzed as subordination. In the analysis that follows, we will suggest that both the Scandinavian and the Semitic constructions should be analyzed as coordination. One might object that the phenomenon in question could not possibly be coordination. Would not typical coordination require the same features in all its conjuncts? Johannessen (1996 and 1998) shows that coordination structures can

 

Coordinated clause structures involving a finite verb form and an infinitive

499

accept different grammatical properties in the conjuncts. Some examples are provided below. In Eastern Mari, a Uralic language spoken in the Mari Republic of the Russian Federation, the expected case-marking is usually only on the second conjunct, as illustrated in (24). (24) Coordination in Eastern Mari Mәj Annan [ydәržö den ergәžәm] palem I Anna.GEN daughter.3SG.POSS and son.3SG.POSS.ACC know.1SG ’I know Anna's daughter and son.’ (Johannessen 1998: 11) Verbal coordination in Japanese cannot take place with an overt conjunction (Bleiler 1978: 97). One way of expressing coordination overtly is exemplified below; the verb of the first conjunct is in a participle form (25): (25) Verbal coordination in Japanese Ginza e [itte miyagemono o okaimasho] Ginza to going souvenirs

OBJ let.us.buy ’Let's go to the Ginza and buy some souvenirs.’ (Bleiler 1978: 97, reproduced in Johannessen 1998: 36)

In Fulfulde, a language in the West-Atlantic branch of the Niger-Kordofanian languages, coordination of VPs can be expressed by serial verb-forms (SI). In this situation, it is possible for the first conjunct (no subsequent ones) to have the ordinary form (here: habitual, HAB) that also occurs in non-coordinated structures (26): (26) Verbal coordination in Fulfulde Janngo Tomorrow

mi yahay, mi foonda ki, mi ndaara ... [I go.SG.HAB I investigate.SG.SI it I see.SI] ’Tomorrow I shall go and investigate it and see ...’ (Arnott 1970, quoted in Endresen 1995, reproduced in Johannessen 1998: 35) Verbal coordination in Modern Turkish usually occurs without an overt conjunction, so that coordination is expressed by juxtaposition. The first con-

 

500

Janne Bondi Johannessen and Lutz Edzard

junct may lack markers for past tense and person and number, and may instead get a dependent marker -ip, as in (27): (27) Verbal coordination in Modern Turkish [Kalk-ip gittik] rise-CVB we.went ’We rose and we went.’ (Lewis 1967: 178, reproduced in Johannessen 1998: 36) Johannessen shows that there is a universal tendency for such cases, which she dubs ”unbalanced coordination”. Importantly, it is the specifier of the coordination phrase that has the expected features, while the complement(s) of the phrase can have either unexpected features or even lack of features. Further, given this analysis, the conjunct with the expected features will be the first one in the coordination phrase in SVO languages and the last one in SOV languages, or more generally, in head-complement vs. complement-head languages. The four languages illustrated in (24) – (27) all have this correlation. Eastern Mari, Japanese and Turkish are OV languages, while Fulfulde is a VO language. The reason for this correlation is that the conjunction is head of the coordination phrase. Since specifiers share their features with the head (specifier-head agreement), and hence that the coordination phrase as such has the relevant features of its specifier, the specifier of the conjunction phrase is the one that interacts with the rest of the clause, and therefore must have the appropriate features (tense, aspect, case, agreement). The complement of the conjunction (the other conjunct) is exempt from this kind of requirements, and the features or lack of features of this part of the conjunction phrase is often language-specific or category-specific. It is expected, within the same language, that the order of head and complement is the same whatever the category of the head, which is why the complement of the conjunction is expected to be the same as that of other heads and complements in the same language (such as verb-object, adposition-complement). Thus, the complement of the conjunction will be preposed to the conjunction in a general OV language and postposed to it in a VO language. The VO conjunction phrase is illustrated in (28a) and the OV one in (28b):

Coordinated clause structures involving a finite verb form and an infinitive

501

(28) Syntactic analysis of VO (a) and OV (b) conjunction phrases (a) (b) a.

b.

CoP[X] X

first conjunct Co conjunction

CoP[X] Co'

Co' Y second conjunct

Y

X Co second conjunct

first conjunct conjunction

The languages that we focus on in this paper are Scandinavian ones (mainly Norwegian, Swedish and Danish), all of which have VO word order, and selected Semitic ones. The latter differ with respect to word order type; while Amharic is SOV, the others are SVO. We illustrate this circumstance with Norwegian, Safaitic, and Amharic examples. The Norwegian example (1) is analyzed here as follows (29): (29) Analysis of pseudo-coordination in Norwegian [og vente]] De ble [CoP stående wait.INFINITIVE they become.PRET stand.PRES.PART and ’They remained standing, waiting.’ (adapted from Lødrup 2002: 138) Here, stående ’standing’ is in the specifier position of the conjunction phrase (CoP), while vente ’wait’ is in the complement position. The whole CoP is of a present participle type: CoP[pres.part].  This a result of the conjunction having entered into Spec-Head agreement with the specifier, which is a present participle. The features of the conjunction and the specifier are unified, which give the phrase both the conjunction features from the head and the participal features from the specifier. The complement of the conjunction has no features that are inherited to the CoP level. It is therefore free to live its grammatical life as it pleases. For the record, it does have to be a verb. Johannessen (1998) shows that the complement conjunct must have major features that are in accordance with the rest of the sentence outside the CoP level. It therefore has to have verbal features. The infinitive is chosen since it is the default verb form in Norwegian, and the one that’s used in dictionary entries. Turning now to Semitic, we start by an example from a SVO language, Safaitic, repeated from (21):

 

502

Janne Bondi Johannessen and Lutz Edzard

(30) Analysis of pseudo-coordination in Safaitic h-rmḫ bql [w [CoP rʿy DEF-camel.COL herbage CONJ pasture.PRF.3.SG.M

kmʾt]] gather_truffles.INF

‘He pastured the camels on spring herbage and gathered truffles.’ (MAHB 5) As in Norwegian, the specifier is the first conjunct and the complement the last conjunct. Because of specifier-head agreement the full conjunction phrase is CoP[PRF], where the PRF features of the specifier and the head have been unified. The complement has no direct way of influencing the CoP, and hence its features are not inherited by the CoP. Instead, it can be tentatively said to have a default verbal form, the infinitive. We then proceed to analyze the Amharic example in (31), repeated from above (23): (31) Analysis of converb coordination in Amharic [täqämmäṭä]] [CoP gäbto sit_down.PRF.3SG.M come_in.CVB.3SG.M ’He came in and sat down.’ (“[he] coming in he sat down”) Since Amharic is a SOV language, here the specifier must be the second conjunct, while the complement is the first conjunct. This time, as expected, it is the second conjunct that has the normal finite features that the verb of the sentence should have, while the first conjunct, the complement, has no direct contact with the rest of the sentence, and only needs to be verbal. The converb features are suitable for this choice. One might ask how there can be complements and specifiers when there is no head and no conjunction. However, as argued by Johannessen (1998: 84-90), asyndetic coordination does have a head, i.e. a conjunction. There are several reasons for this assumption: 1) In some languages (e.g. Sissala, a Niger Congo Voltaic language, and Sarcee, an Athapaskan language spoken in Alberta) conjunctions are facultative. 2) Many languages lack a conjunction for certain categories and phrases, but not for others (cf. the Japanese example in (25) above). 3) Languages that lack an overt conjunction in conjunctive coordination, still have one in disjunctive coordination (this is the case in many Turkic languages). 4) Some languages, like Turkish, have borrowed conjunctions from other languages. Nothing else is changed in the overall structure, which strongly sug-

 

Coordinated clause structures involving a finite verb form and an infinitive

503

gests that there has been a zero conjunction before that has been substituted with an overt one. 5) Child language often lacks conjunctions even in languages that otherwise have them. To sum up, in this section we have shown that a coordination analysis as that proposed in Johannessen (1997, 1998) is well suited for pseudo-coordination and converb constructions in Scandinavian and Semitic languages, respectively. When the verbal complexes are analyzed as coordination, one verb is taken to be specifier and one to be complement in the conjunction phrase. From this it follows that if one of them has deviant morpho-syntactic features compared with what one would have expected, it is the verb in the specifier position that will be normal and the complement verb that will be deviant. It is the overall word order type of the language that shows which of the conjuncts is specifier or complement. We have shown that the Scandinavian VO language Norwegian and the Semitic VO language Safaitic have the deviant verb, the second conjunct, in the complement of the conjunction phrase, while the Semitic OV language Amharic has its deviant verb in the first conjunct, which is the complement of the conjunction phrase in this language. 5 Conclusion We hope to have shown that it is not only interesting but indeed meaningful to compare seemingly idiosyncratic structures, here a finite verb form coordinated with an infinitive, in two unrelated language families. Finding typological parallels has a clear potential of shedding light on constructions that are otherwise hard to understand.

References Al-Jallad, Ahmad. 2015. An Outline of the Grammar of the Safaitic Inscriptions. Leiden: Brill. Appleyard, David. 1995. Colloquial Amharic: A complete language course. London: Routledge. Arnott, David W. 1970. The Nominal and Verbal System of Fula. Oxford: Oxford University Press. Bleiler, Everett F. 1978. Basic Japanese Grammar. Tokyo: Charles E. Tuttle Company. Brockelmann, Carl. 1913. Grundriß der vergleichenden Grammatik der semitischen Sprachen. II. Band: Syntax. Berlin: Verlag von Reuther & Reichard.

 

504

Janne Bondi Johannessen and Lutz Edzard

Edzard, Lutz. 2014. “The finite-infinite dichotomy in a comparative Semitic perspective.” In: Ronny Meyer, Yvonne Treis, and Azeb Amha (eds.). Explorations in Ethiopian Linguistics: Complex Predicates, Finiteness, and Interrogativity, 205–233. Wiesbaden: Harrassowitz. Endresen, Rolf T. 1995. “Norwegian og and å – a cognitive view.” Nordic Journal of Linguistics 18: 201–218. Faarlund, Jan Terje, Svein Lie, and Kjell Ivar Vannebo. 1997. Norsk referansegrammatikk. Oslo: Universitetsforlaget. Hackett, Jo Ann. 2013. Presentation on Phoenician inscriptions at the University of Oslo, May 24, 2013. Hansen, Erik and Lars Heltoft. 2011. Grammatik over det Danske Sprog. Det Danske Sprog- og Litteraturselskab. Odense: Syddansk Universitet. Hetzron, Robert. 1972. Ethiopian Semitic. Studies in Classification. Manchester: Manchester University Press. Huehnergard, John. 2005 (2nd ed.). A Grammar of Akkadian. Winona Lake: Eisenbrauns. Huesman, John. 1956. ”Finite uses of the infinitive absolute.” Biblica 37: 271–295. Johannessen, Janne Bondi. 1996. ”Partial agreement and coordination. Remarks and Replies.” Linguistic Inquiry 27: 661–676. Johannessen, Janne Bondi. 1998. Coordination. Oxford University Press, New York, Oxford. KRS = Safaitic inscriptions and drawings recorded by the Basalt Desert Rescue Service and to be published by Dr G. M. H. King. Lewis, Geoffrey. 1967. Turkish Grammar. Oxford: Clarendon Press. Lipiński, Edward. 2001 (2nd ed.). Semitic Languages. Outline of a Comparative Grammar. Leuven: Peeters. Lipiński, Edward. 2010. “Le gérondif en phénicien”, Journal of Semitic Studies 55/1: 1–10. Lødrup, Helge. 2002. “The syntactic structures of Norwegian pseudocoordinations.” Studia Linguistica 56/2: 121-143. Lødrup, Helge. 2014. ”There is no reanalysis in Norwegian pseudocoordinations (except when there is).” In: Hans Petter Helland and Christine Meklenborg Salvesen (eds.) Affaire(s) de grammaire, 43–65. Oslo: Novus Press. Merve, Christo H.J. and Alexander Andrason. 2014. “Finite infinitive? ’Finite’ uses of the Biblical Hebrew infinitive absolute and their rationale.” Journal of Semitic Studies 59/1: 255–296. Meyer, Ronny. 2012. “The converb in Amharic.” In: Domenyk Eades (ed.). Grammaticalization in Semitic, 165–191. Oxford: Oxford University Press. Morrison, Craig E. 2013. “Infinitive: Biblical Hebrew.” In: Geoffrey Khan (ed.). Encyclopedia of Hebrew Language and Linguistics, Volume 2: G–O, 266–269. Leiden: Brill. Rubin, Aaron D. 2005. Studies in Semitic Grammaticalization. Winona Lake, Indiana: Eisenbrauns.

 

Coordinated clause structures involving a finite verb form and an infinitive

505

Rubinstein, Arie. 1952. ”A finite verb continued by an infinitive absolute in Hebrew.” Vetus Testamentum 2: 362–367. Stowell, Tim. 1982. ”The tense of infinitives.” Linguistic Inquiry 13: 561–570. Waltke, Bruce K. and Michael O’Connor. An Introduction to Biblical Hebrew Syntax. Winona Lake: Eisenbrauns. Wiklund, Anna-Lena. 2007. The Syntax of Tenselessness Tense/Mood/Aspect-agreeing Infinitivals. Berlin/New York: Mouton de Gruyter. Yri, Kjell Magne. 2014. “Finiteness as manifested by grounding and deixis: Amharic (Semitic) and Sidaama (Cushitic).” In: Lutz Edzard and John Huehnergard (eds.). Proceedings of the Oslo–Austin Workshop in Semitic Linguistics. Oslo, May 23 and 24, 2013, 95–126. Wiesbaden: Harrassowitz.

 

Causative constructions in Neo-Aramaic (Christian Urmi dialect) Geoffrey Khan, Cambridge University

Semitic causative verbal forms were a central concern of Jan Retsö’s groundbreaking work Diathesis in the Semitic Languages (1989). It is my great pleasure to offer here a small contribution to my friend and colleague about the causative constructions of Neo-Aramaic. The presentation is based on material I have gathered on the spoken dialect of the Assyrian Christians of Urmi (northwestern Iran), a North-Eastern Neo-Aramaic (NENA) dialect. This material has been collected from native speakers originating from the town of Urmi and various villages in the Urmi area.1 Causatives are regarded by some linguists as expressing complex events that can be decomposed into a causing microevent and a caused microevent with a dependency relationship between the two (Shibatani 1976; Comrie 1981, 158–177; Kulikov 2001; Song 2006). The semantic structure of causatives can be represented as follows (Van Valin 2006):2 CAUSE (x, [EVENT (y, (z))]) where x is the subject of the causative verb, y and z are participants in the caused event. There may be different numbers of participants in the caused event according to the lexical verb. Participant y corresponds to the subject of the non-causative verb and participant z corresponds to its object. Dixon (2000: 30) prefers to characterize a causative as an increase in valency involving ‘the specification of an additional argument, a causer, onto a basic clause’. There are three basic types of causative: (i) lexical causative, (ii) morphological causative and (iii) periphrastic causative.

1

The full description of the dialect will appear shortly in Khan (forthcoming). For a description of causative constructions in other NENA dialects see Göransson 2014.

2

This formalism derives ultimately from the decompositional system of Dowty 1979.

Causative constructions in Neo-Aramaic (Christian Urmi dialect

507

A lexical causative expresses the addition of the causer argument by a lexical item rather than a modifcation of morphology. Lexical causatives are of two kinds: (a) when there are two unrelated lexical items that appear to be in a causative relation and (b) when a single lexeme can be used in either a causative or a non-causative function (Dixon 2000, 38). An example cited for (a) in English is kill, which may be considered to be the causative of die. Category (b) of lexical causatives include ambitransitive ‘labile’ verbs that can be used intransitively or transitively, the subject of the intransitive alternant being the object of the transitive alternant, e.g. English melt, trip. The transitive alternant can be considered to be a causative of the intransitive alternant. A morphological causative is expressed by a distinctive morphological process, such as affixation or internal patterning. A periphrastic causative is expressed by two verbs, one denoting the causing and the other the lexical content of the event. These different typologies of causative exhibit varying degrees of formal compactness in encoding the causative event. Lexical causatives In the Christian Urmi Neo-Aramaic dialect (henceforth C. Urmi) there is a dearth of lexical causatives consisting of different lexical items. One example is myt ‘to die’ and +k̭ṱl ‘to kill’, as in English. There are, however, a series of labile verbs consisting of intransitive and transitive alternants. In such cases the subject of the intransitive alternant is the affectee of the event expressed by the verb and does not actively initiate it. The affectee subject undergoes a change of state and is equivalent semantically to the object (accusative) of a transitive verb, and so may be termed ‘unaccusative’ subject. A feature of verbs of this category is that they do not grammatically encode an agent or cause but rather are encoded as spontaneous, or ‘inchoative’ according to the terminology of Haspelmath (1993). C. Urmi has triliteral and quadriliteral verbal roots. Triliteral roots have three verbal stems corresponding to the pәʿal (stem I), the paʿʿel (stem II) and the ʾap̄ʿel (stem III) forms of earlier Aramaic. The vast majority of quadriliteral verbs have only one basic stem (Q I). A few quadriliterals have, in addition, a derived stem based on stem III of triliteral verbs (Q II). Labile verbs consist of intransitive and transitive alternants in the same verbal stem, without any change of morphological form. The vast majority of these

508

Geoffrey Khan

verbs are triliteral verbs of stem I or quadriliteral verbs of the basic Q I stem. Examples of triliteral stem I verbs include: ptx I ‘to open’ (tr. and intr.) bәt-patxi (1) +tarranә FUT-open.3PL doors ‘The doors will open’ (unaccusative intransitive) (2) poxaMS

+

tarranә bәt-patәx-lun FUT-CAUSE.open.3MS-OBJ.3PL wind doors ‘The wind will open the doors’ (transitive)

+

šmṱ I ‘to break’ (tr. and intr.)

bәt-+šamṱi stones FUT-break.3PL ‘The stones will break’ (unaccusative intransitive) cipә bәt-+šamәṱ-lun (4) čacučM   hammer stones FUT-CAUSE.break.3MS-OBJ.3PL ‘The hammer will break the stones’ (transitive)

(3) cipә

The events expressed by the intransitive alternants of such labile verbs result in a change of configurational state that is conceptualized as typically coming about through the effect of an external cause, although in the intransitive alternants the event is presented as spontaneous without the external cause being grammatically encoded. In the transitive alternant forms, on the other hand, the cause is grammatically encoded. Intransitive unaccusative verbs that are not labile include verbs expressing events that have an internal cause, although not grammatically encoded, e.g. pc̭x I ‘to blossom’, +č̭mč̭ I ‘to wither’, +ṱrs I ‘to become fat’,3 verbs expressing events that are conceptualized as typically spontaneous, e.g. brz I ‘to become dry’, smk̭ I ‘to become red’, pšr I ‘to melt’. Intransitive verbs of movement, which are generally considered to have unaccusative subjects (Levin and Rappaport Hovav 1995), are, likewise, never labile. Many quadriliteral verbs in the basic Q I stem are labile in that they can be used as unaccusative intransitives or transitives without changing their morphological form, e.g.

3

For the notion of internal cause see Alexiadou, Anagnostopoulou, and Schäfer 2006.

Causative constructions in Neo-Aramaic (Christian Urmi dialect

509

brbz Q I ‘to scatter’ (intr. and tr.) bәt-barbәzzi ɟu-+ʾatra people FUT-scatter.3PL in-land ‘The people will scatter in the land’ (unaccusative intransitive)

(5) našә

(6) malca bәt-barbәz-lun

našә

kingMS FUT-CAUSE.scatter.3MS-OBJ.3PL people ‘The king will scatter the people’ (transitive) jnɟn ‘to rust; to cause to rust’ (7) prәzla bәt-janɟәn ironMS FUT-rust.3MS ‘The iron will rust’ (unaccusative intransitive) prәzla (8) +mәṱra bәt-janɟәnna-lә rainFS FUT- CAUSE.rust.3FS-OBJ.3MS iron ‘The rain will cause the iron to rust’ (transitive) Morphological causatives Causatives of stem I verbs are most frequently expressed morphologically by stem III (the historical ʾap̄ʿel form). In some cases a causative is expressed by stem II (the historical paʿʿel form), but the formation of causatives by stem II is not as productive or transparent as it is by stem III. In C. Urmi all inflected forms of stem III have a /m/ prefix. This originated in the participles, but has been extended by analogy to all verbal bases. Following Goldenberg (2005: 16; 2013: 127), the /m/ can be considered an augment to the radicals of the root: Stem I: Stem III

root: root

brz ‘to dry’ m-brz ‘to cause to dry’

Patterns of verbal bases of stem III: Present base Past base

maC1C2әC3 muC1C2әC3

Resultative participle muC1C2әC3C3a Imperative maC1C2әC3 Infinitive maC1C2uC3ә Verbal noun maC1C2aC3ta

mabrәz mubrәz mubrәzza mabrәz mabruzә mabrazta

510

Geoffrey Khan

These consist of discontinuous morphemic patterns that are merged with the radicals of the verbal root. The radicals of the root are anchored to three slots. In some cases the slots contain a cluster of two radicals: Radical slot

1

2

3

Present base

m a

C1C2 ә

C3

Past base

m u C1C2 ә

C3

Resulative participle m u C1C2 ә

C3C3 a

Infinitive

m a

C1C2 u C3

ә

Verbal noun

m a

C1C2 a C3

ta

There is marginal use of a morphological causative of quadriliteral verbs with a prefixed /m/, e.g. Stem Q I: root: Stem Q II root

dmdm m-dmdm

‘to bleed’ ‘to cause to bleed’

Patterns of verbal bases of stem Q II Present base mC1aC2C3әC4 Past base mC1u C2C3әC4 Resultative participle mC1uC2C3әC4C4a Imperative mC1aC2C3әC4 Infinitive mC1aC2C3uC4ә Verbal noun mC1aC2C3aC4ta

mdamdәm mdumdәm mdumdәmma mdamdәm mdamdumә mdamdamta

This exhibits the same patterns as stem III of triliteral verbs, but with a cluster of two radical consonants being inserted into the first of the three slots: Radical slot

1

2

3

Present base

mC1 a

C2C3 ә

C4

Past base

mC1 u C2C3 ә

C4

Resulative participle mC1 u C2C3 ә

C4C4 a

Infinitive

mC1 a

C2C3 u C4

ә

Verbal noun

mC1 a

C2C3 a C4

ta

Patterns of stem II in bases of triliteral verbs are as follows:

Causative constructions in Neo-Aramaic (Christian Urmi dialect

511

spk̭ II ‘to empty (tr.)’ sapәk̭ supәk̭ supk̭a

Present base

C1aC2әC3 Past base C1uC2әC3 Resultative participle C1uC2C3a Imperative C1aC2әC3 Infinitive C1aC2uC3ә Verbal noun C1aC2aC3ta

sapәk̭ sapuk̭ә sapak̭ta

The vocalic patterns of stem II inflections are almost identical to those of stem III. The formal differentiation consists of the lack of root augment and the different distribution of the root consonants in the 3 slot template. Due to the lack of root augment, stem II can be considered to be morphologically less complex than stem III: Radical slot

1

2

3

Present base

C1 a

C2 ә

C3

Past base

C1 u C2 ә

C3

Resulative participle C1 u C2 — C3 A Infinitive

C1 a

C2 u

C3 ә

Verbal noun

C1 a

C2 a

C3 Ta

The only difference between stem II and stem III with regard to vocalic pattern is in the resultative participle. This is eliminated, however, if we consider the stem III pattern of the participle muC1C2әC3C3a to have originally been muC1C2C3a and the ә between the second and third radicals to have been inserted as an epenthetic with consequential gemination of the third radical. This was undoubtedly the original phonological representation of the pattern, but in the synchronic state of the dialect, it would appear that the ә has become phonologized. Morphological causatives of triliteral verbs are formed by stem II from a number of stem I intransitives. The intransitives in all cases are inchoative with unaccusative subjects, e.g. +

bsr I plṱ I

‘to diminish’ ‘go out’

+

trs I bšl I bsm I

‘to mend (intr.)’ ‘to cook (intr.)’ ‘to become well’

+

+ +

bsr II plṱ II

‘to reduce (tr.)’ ‘to put out’

trs II bšl II bsm II

‘to repair’ ‘to cook (tr.)’ ‘to cure’

+

512

Geoffrey Khan

k̭ly I

‘to become parched’

k̭ly II

‘to parch’

pny I pšr I spk̭ I špx I

‘to be exterminated’ ‘to melt (intr.)’ ‘to empty (intr.)’ ‘to pour out (intr.)’

pny II pšr II spk̭ II špx II

‘to exterminate’ ‘to melt (tr.)’ ‘to empty (tr.)’ ‘to pour out (tr.)’

The causee, i.e. the grammatical subject of the intransitive verb, is made the grammatical object of the causative verb, e.g. + ʾal-+ʾarra (9) miyya bәt-šapxi water FUT-pour.3PL on-ground ‘The water will pour out on the ground’

miyya (10) +ʾav bәt-šapәx-lun FUT-CAUSE.pour.3MS-OBJ.3PL waterPL he ‘He will pour out the water on the ground’

+

ʾal-+ʾarra on-ground

The relationship between such pairs of stem I and stem II forms is similar to that holding between the intransitive and transitive alternants of stem I labile verbs, in that in labile verbs the intransitive is always unaccusative inchoative. Stem II is not used to derive causatives from agentive intransitives, or verbs of higher valencies (transitives or ditransitives). Stem III morphological causatives of triliteral verbs are more productive than stem II causatives. They are used to derive causatives of both unaccusative and agentive stem I intransitive verbs, and also causatives of transitive verbs. Stem II causatives, which are less productive than Stem III causatives, tend to express more common unmarked, i.e. conventional, situations (e.g. ‘to cook’, ‘to cure’) than stem III. Unaccusative stem I

Causative stem III

brz ‘to dry’ cry ‘to become short’

m-brz ‘to cause to dry’ m-cry ‘to cause to become short’

Agentive stem I

Causative stem III

brc ‘to kneel’

m-brc ‘to cause to kneel’

Intransitive verbs with agentive subjects are sometimes termed ‘unergative’ (Perlmutter 1978). As is the case with stem II, when intransitive verbs are made into causatives by stem III, the causee, i.e. the grammatical subject of the intransitive verb, is made the grammatical object of the causative verb. This applies to both unaccusative subjects and agentive subjects of stem I verbs, e.g.

Causative constructions in Neo-Aramaic (Christian Urmi dialect

513

m-brz ‘to cause to dry’ (11) +ʾav

bәt-mabrәz-lun FUT-CAUSE.dry.3MS-OBJ.3PL he ‘He will cause the wheat to dry’

+

xәṱṱә wheat

m-brc ‘to cause to kneel’ bәt-mabrәc-lun našә (12) +av FUT-CAUSE.kneel.3MS-OBJ.3PL people he ‘He will make the people kneel’ This syntax, whereby the subject of an intransitive is made the object of a causative verb, is the cross-linguistic norm. In many languages morphological causatives can be formed only from intransitive verbs (Dixon 2000: 45). The application of morphological causatives to transitive verbs with two arguments (subject and direct object) is less common cross-linguistically. When this occurs, moreover, various syntactic alignments of the arguments are attested in documented languages. These are summarized by Dixon (2000: 48) thus (where A = transitive subject, and O = object): type

causer original A (causee)

original O

(i) (ii)

A A

special marking retains A-marking

O O

(iii) (iv) (v)

A A A

has O-marking O non-core

has O-marking non-core O

The stem III morphological causatives of transitive verbs in C. Urmi exhibit type (v) alignment, in that the original object of the transitive clause remains the object of the causative verb, but the original subject of the transitive clause is expressed by an adjunct prepositional phrase. When the subject of the original clause is agentive, the preposition of the adjunct is normally b- ‘by’, e.g. Stem I transitive mxy ‘to hit’ (13) ʾo naša bәt-maxi-lun that man FUT-hit.3MS-OBJ.3PL ‘That man will hit the dogs’

calbә dogs

Stem III causative m-mxy ‘to cause to be hit’ (14) bәt-mamxi-lun calbә b-do FUT-CAUSE.hit.3MS-OBJ.3PL dogs by-OBL.that ‘He will cause the dogs to be hit by that man’

naša man

514

Geoffrey Khan

On account of this alignment of the arguments of morphological causatives, it is more appropriate to translate a causative verb such as m-mxy ‘to cause to be hit’ rather than ‘to cause to hit’. If the subject of the transitive verb is an experiencer, the preposition of the adjunct is normally the dative preposition k̭a- ‘to’, e.g. Stem I transitive xzy ‘to see’ (15) bab-u bәt-xazә ctava father-his FUT-see.3MS book ‘His father will see a book’ Stem III causative m-xzy ‘to cause to be seen’ (16) bәt-maxzә ctava k̭a-bab-u FUT-CAUSE.see.3MS book to-father-3MS ‘He will cause a book to be seen by his father’ (i.e. ‘He will show a book to his father’) Stem I transitive +šmy ‘to hear’ bәt-+šamә hak̭yat (17) +xor-u friend-his FUT-hear.3MS story ‘His friend will hear a story’ Stem III causative +m-šmy ‘to cause to be heard’ bәt-+mašmә hak̭yat k̭a-+xor-u FUT-CAUSE.hear.3MS story to-friend-his he ‘He will cause a story to be heard by his friend’.

(18) +ʾav

Comrie (1976) proposed that when causatives of transitives follow a type (v) alignment the causee, ousted from the subject position by the causer, is demoted to the highest (= leftmost) free position down a grammatical relations hierarchy, also known as ‘noun phrase accessibility hierarchy’, with the following form: Subject > Direct object > Indirect object > Oblique object C. Urmi constructions with experiential causees would conform to this principle, in that dative k̭a-phrases correspond formally to indirect objects. If the causee is non-experiential, however, it skips one place in the hierarchy since it is expressed as an oblique object in a b-phrase, which corresponds to the agentive adjunct of a passive construction. These differing degrees of demotion down the hierarchy of experiential and non-experiential causees can be correlated with the fact that experiential causees, although agents, are affected,

Causative constructions in Neo-Aramaic (Christian Urmi dialect

515

whereas non-experientail causees are unaffected agents (Saksena 1980; Kulikov 2001: 890–891). Stem III may be used to produce causatives of the transitive alternants of labile stem I verbs, e.g. ptx ‘to open’ (tr. and intr.). These causatives have the same syntactic alignment with the causee expressed in an adjunct phrase. Stem I labile ptx ‘to open (intr. and tr.)’ (19) +tarranә bәt-patxi FUT-open.3PL doors ‘The doors will open’ (unaccusative intransitive) bәt-patәx-lun (20) ʾo naša +tarranә FUT-CAUSE.open.3MS-OBJ.3PL that man doors ‘That man will open the doors’ (transitive) Stem III causative m-ptx ‘to cause to be opened’ bәt-maptәxx-әn-lun b-do (21) ʾana +tarranә FUT-CAUSE.open-1MS-OBJ.3PL by-OBL.that I doors ‘I shall cause the doors to be opened by that man’ (causative)

naša man

If the verb is unaccusative intransitive in stem I and expresses a causative of the intransitive by stem II, then, likewise, the stem III form expresses a causative of the transitive: prk̭ I prk̭ II m-prk̭ III

‘to finish’ (intr.) ‘to finish (tr.) ‘to cause to be finished by so.’

FINISH (y) CAUSE (x[FINISH (y)]) CAUSE (x[FINISH (y, (z))])

If stem I of a verb expresses an intransitive and there is no labile transitive alternant or transitive stem II to express the causative of the intransitive, then stem III of the verb expresses the causative either of the intransitive or of the corresponding transitive, e.g. cly I m-cly III m-cly III

‘to stand, to stop’ (intr.) ‘to stop’ (tr.) ‘to cause to be stopped by so’

STOP (y) CAUSE (x[STOP (y)]) CAUSE (x[STOP (y, (z))])

The causative Q II form of quadriliteral verbs is used to derive causatives of intransitive verbs, transitive verbs and the transitive alternant of labile verbs. As with stem III causatives of transitives verbs, the object of the original verb remains the object and the causee subject is put in a peripheral position expressed by a prepositional phrase, e.g.

516

Geoffrey Khan

xšcn ‘to become dark’, m-xšcn ‘to cause to become dark’ (22) lelә bәt-xašcәn night FUT-darken.3MS ‘The night will become dark’ (intransitive unaccusative) lelә (23) +ʾayva bәt-mxašcәn-lә cloud FUT-CAUSE.darken-OBJ.3MS night ‘The cloud will make the night dark’ (causative) +

k̭ṱk̭ṱ ‘to chop’, +m-k̭ṱk̭ṱ ‘to cause to be chopped.’

(24) +busra bәt-+k̭aṱk̭әṱṱi-lә meat FUT-cut.3PL-OBJ.3MS ‘They will chop the meat’ (transitive) busra bәt-+mk̭aṱk̭әṱṱi-lә b-do naša meat FUT-CAUSE.cut.3PL-OBJ.3MS by-OBL.that man ‘They will cause the meat to be chopped by that man’ (causative) +

ɟndl Q I ‘to roll’ (intr. and tr.), m-ɟndl Q II ‘to cause to be rolled’

(25) ɟutta bәt-ɟandәlla ball FUT-roll.3FS ‘The ball will roll’ (unaccusative transitive) ɟutta

bәt-ɟandәlli-la FUT-CAUSE.roll.3PL-OBJ.3FS ball ‘They will roll the ball’ (transitive) bәt-mɟandәlli-la ɟutta b-do naša FUT-CAUSE.roll.3pl-OBJ.3FS ball by-OBL.that man ‘They will cause the ball to be rolled by that man’ (causative) The causee subject of a transitive verb can be left unspecified in morphological causative constructions, e.g. (26) bәt-mamxә našә FUT-CAUSE.hit.3MS people ‘He will cause people to be hit’ (27) bәt-maxzә ctava FUT-CAUSE.see.3MS book ‘He will cause a book to be seen (= show a book)’

Causative constructions in Neo-Aramaic (Christian Urmi dialect

517

(28) bәt-+mašmә hak̭yat FUT-CAUSE.hear.3MS story ‘He will cause a story to be heard’ A further variation of structure is a reflexive causative in which one of the arguments of the underlying transitive verb is coreferential with the subject of the causative verb. This is found in the experiential causative m-xzy ‘to cause to be seen’. The coreferential item is the reflexive direct object in the semantic structure, although not expressed on the surface, and the other participant must be expressed by a prepositional adjunct, e.g. k̭a-našә (29) bәt-maxzә FUT-CAUSE.see.3MS to-people CAUSE (xi[SEE (yi)])

‘He will cause himself to be seen by people’ (= ‘He will show himself to people’) This reflexive semantic structure of causatives is found also where the corresponding stem I verb has an unaccusative subject, e.g. m-ɟrp III, the causative of ɟrp ‘to slip’: (30) SLIP (y) bәt-jarәp FUT-slip.3MS ‘He will slip’ (31) CAUSE (xi[SLIP (y)]) bәt-majrәp našә FUT-CAUSE.slip.3MS people ‘He will cause people to slip’ (32) CAUSE (xi[SLIP (yi)]) bәt-majrәp FUT-CAUSE.slip.3MS

‘He will cause himself to slip’ = ‘He will skate’ When a stem I verb has a reflexive semantic structure, such as lvš ‘to dress (oneself)’, in which the subject is both agent and affectee, the object of a causative stem III from the same root can take as its direct object the item that is subject of the stem I verb. This is permitted since the subject is the affectee and semantically equivalent to an object, e.g.

518

Geoffrey Khan

(33) +časibә

bәt-loši paupers FUT-dress.3PL ‘Paupers will get dressed’

bәt-loši jullә (34) +časibә paupers FUT-dress.3PL clothes ‘Paupers will get dressed in clothes’ + casibә (35) bәt-malvәš FUT-CAUSE.dress.3MS paupers ‘He will dress paupers’ + (36) bәt-malvәš casibә jullә FUT-CAUSE.dress.3MS paupers clothes ‘He will dress paupers in clothes’

The argument denoting the clothing should be regarded as a form of adverbial complement rather than a core argument. The same applies to verbs of consumption, such as +ʾxl ‘to eat’ and šty ‘to drink’. The subject of the stem I verb, which is both the agent and the affectee of the consumption, can be made the direct object of the corresponding stem III causative, e.g. bәt-+ʾaxli (37) +raba našә many people FUT-eat.3PL ‘Many people will eat rice’

+

rәzza

rice

+ + raba našә rәzza (38) bәt-+maxә FUT-CAUSE.eat.3MS many people rice ‘He will cause many people to eat rice’ (= He will feed many people rice’)

A similar semantic structure is exhibited by the transitive verb +k̭ry ‘to study’, in which the subject is also the affectee of the activity, in that he becomes ‘learned’. The causative stem III form +m-k̭ry is used thus: + raba mәndyanә (39) bәt-+k̭ar-әt FUT-study-2MS many things ‘You will study many things’ + raba mәndyanә ʾana bәt-+mak̭r-әn-lux FUT-CAUSE.study-1MS-OBJ.2MS many things I ‘I shall cause you to study many things (= I shall cause you to be educated in many things)’

Causative constructions in Neo-Aramaic (Christian Urmi dialect

519

The verb xyp I ‘to wash’ may be used intransitively with the semantic structure of an agentive reflexive, in that the subject is agentive but the referent of the subject is also the affectee of the action, though not expressed: (40) bәt-xayәp FUT-wash.3MS ‘He will wash (himself)’ The verb may also be used as a transitive agentive with an explicit object: (41) bәt-xep-a-la brat-i FUT-wash-3FS-OBJ.3FS daughter-my ‘She will wash my daughter’ The stem III form m-xyp is used to express the causative of the agentive transitive, with the affectee as the grammatical object and the causee agent optionally expressed in an adjunct phrase: (42) bәt-maxyәp-la brat-u b-de FUT-CAUSE.wash.3MS-OBJ.3FS daughter-his by-OBL.that ‘He will cause his daughter to be washed by that woman’

baxta woman

The same applies to the verb šlx I ‘to undress’, which may be used as a agentive reflexive or an agentive transitive. When used in the causative stem III form, the object of the verb is the affectee of the agentive transitive and the causee agent is expressed by an optional adjunct phrase, e.g. (43) bәt-mašlәx-lә b-do FUT-CAUSE.undress.3MS-OBJECT.3MS by-OBL.that ‘He will cause him to be undressed by that man’

naša man

So-called ‘inner objects’ of verbs behave differently from regular objects in causative constructions. An ‘inner object’ is typically a verbal noun or abstract derived from the same root as the verb. Unlike a regular object, an inner object is not a pre-existing entity that is affected by the action of the verb but rather comes about when the action takes place. In fact the verbal action and the inner object are one and the same thing (Ilani 2013), e.g. bәt-zamәr zmarta FUT-sing.3MS song ‘He will sing a song.’ This is shown by the fact that an inner object nominal can occur with intransitive verbs that do not have a patient object slot in their semantic structure, e.g.

520

Geoffrey Khan

(44) xi-lә xayyә +yarixә LIVE.past-ERG.3MS life long ‘He lived a long life’ The causative of zmr I ‘to sing’, m-zmr III, may take the subject of the embedded event (i.e. the one who sings) rather than the inner object as the direct object. The inner object may optionally follow the regular object of the causative, e.g. (45) bәt-mazmәr xa naša zmarta FUT-CAUSE.sing.3MS one man song ‘He will cause somebody to sing a song’ Alternatively the inner object can be made the object of the causative verb and the subject put in a prepositional adjuct, as in the regular causative construction, e.g. (46) bәt-mazmәr zmarta b-xa naša FUT-CAUSE.SING.3MS song by-one man ‘I will cause a song to be sung by somebody.’ In some cases a stem I form verb takes a complement in the form of a prepositional phrase, e.g. (47) bәt-xadә b-do mәndi FUT-rejoice.3MS by-OBL.that thing ‘He will be happy with that thing’ + b-do šula (48) bәt-balәɟ FUT-busy.3MS with-OBL.that job ‘He will be busy with that job’

(49) bәt-tapәk̭

b-do

naša

FUT-meet.3MS with-OBL.that man ‘He will meet that man’

b-do (50) bәt-dak̭әr FUT-knock.3MS on-OBL.that ‘He will knock that man’ b-do (51) bәt-k̭ayәt

naša

FUT-touch.3MS on-OBL.that ‘He will touch that man’

man

man naša

In such cases the causative stem III takes the subject of the stem I verb as its object when the causee subject is non-agentive and the referent in the preposi-

Causative constructions in Neo-Aramaic (Christian Urmi dialect

521

tional phrase is not affected by the event. This applies to the causative of (47) – (49) above: (52) bәt-maxd-әn-lә b-do mәndi FUT-CAUSE.rejoice-1MS-OBJ.3MS with-OBL.that thing ‘I shall cause him to be happy with that thing’ (53) bәt-mablәɟɟ-әn-lә

b-do FUT-CAUSE.busy-1MS-OBJ.3MS by-OBL.that ‘I shall cause him to be busy with that job’

+

šula job

b-do naša (54) bәt-matpәk̭k̭-әn-lә FUT-CAUSE.meet-1MS.OBJ.3MS with-OBL.that man ‘I shall cause him to meet that man’ Morphological causatives cannot be derived from constructions such as (50) and (51), in which the subject is agentive and the referent of the prepositional phrase is affected by the event. Although the referent of the prepositional phrase is an affectee, it is not a core object argument but an adjunct. A morphological causative of a transitive verb requires a core object argument, the adjunct slot being taken by the agent. Compare the following: (55) bәt-davәk̭-la FUT-seize.3MS-OBJ.3FS ‘He will seize her’ biyy-o (56) bәt-davәk̭ fut-seize.3MS on-3FS ‘He will hold onto her’ A stem III morphological causative may be formed from (55) but not from (56): biyy-u (57) bәt-madvәk̭k̭-әn-la FUT-CAUSE.seize-1MS-OBJ.3FS by-3MS ‘I shall cause her to be seized by him’ The stem III causative verb +m-šmy can be used in the sense of ‘to listen’, e.g. + ʾal-našә (58) bәt-+mašmә FUT-CAUSE.hear.3MS to-people ‘He will listen to people’

This may be interpreted as a construction with an underlying experiential, i.e. non-agentive, reflexive causee and a prepositional adjunct with a referent that is not affected by the event: ‘He will cause himself to hear to people’. It

522

Geoffrey Khan

would, therefore, be analogous to constructions such as (52) – (54) above. Alternatively it could be interpreted as the use of a morphological causative to express a valency-increasing construction with an oblique argument known as an ‘applicative’ (Dixon 2000: 31; Kulikov 2001: 894). The causative stem III of some verbs has undergone a metaphorical semantic shift, e.g. + +

k̭rm I

‘to win’ m-k̭rm III ‘to cause to be won’ = ‘to lose’

+

k̭rṱ I

+

m-k̭rṱ III

‘to bite, to lacerate’ ‘to cause to be bitten (by scissors)’ = ‘to shear, cut (hair)’

A morphological causative verb cannot be used to form a causative of a ditransitive clause containing two affectee arguments, such as (59) bәt-tanә hak̭yat k̭atux ‘I tell you a story’ A morphological causative verb can, however, form a causative of compound verbal construction that consists of a noun and verb. e.g. +

nala mxy ‘to shoe (a horse)’ (literally: ‘to strike shoe’)

(60) bәt-max-әn-lun FUT-strike-1MS-OBJ.3PL ‘I shall shoe my horses’

suysavat-i +nala horses-1S shoe

(61) bәt-mamx-әn-lun suysavat-i +nala b-do FUT-CAUSE.strike-1MS-3PL horses-1S shoe by-OBL.that ‘I shall cause my horses to be shod by that man’

naša man

This is permitted because the noun that is combined with the verb in such compound constructions is not an affected argument of the verb but rather a unaffected complement. They, therefore, do not form true ditransitive constructions. Periphrastic causatives An alternative means of expressing the causative is the use of a biclausal periphrastic construction with the verb m-∅vd, which is the stem III morphological causative of the verb ∅vd ‘to do’. The causer is made the subject of the verb m∅vd. The caused event is expressed by a following subordinate clause, e.g. + (62) bәt-mavvәdd-әn šar-әt FUT-CAUSE.do-1MS begin-2MS ‘I shall cause you to begin’

Causative constructions in Neo-Aramaic (Christian Urmi dialect

523

The subordinate clause may be optionally introduced by the subordinator ʾәt or k̭at. This demonstrates that the construction is biclausal and the head verb has not fused with the lexical verb and become an auxiliary verb, as is the case in periphrastic causatives in some languages (Kulikov 2001: 893), e.g. k̭at (63) bәt-mavvәddәn FUT-CAUSE.do-1MS that ‘I shall cause you to begin’

+

šar-әt begin-2MS

These constructions are in conformity with the usual syntactic alignment of morphological causatives, in that the verb can be considered to have the meaning ‘cause to be done’ with its object, the subordinate clause, expressing what is done. On many occasions, however, the subject argument of the subordinate clause, i.e. the causee, is raised into the direct object position of the verb m-∅vd, e.g. (64) mavvәdd-әn-lux FUT-CAUSE.do-1MS-OBJ.2MS ‘I shall cause you to begin’

k̭at

+

that

begin-2MS

šar-әt

In periphrastic causative constructions the original syntactic alignment of the caused clause is maintained in the subordinate clause. The subject of both intranstive and transitive clause remains the subject of the subordinate clause, e.g. (65) +ṱerә bәt-parx-i birds FUT-fly-3PL ‘The birds will fly’ (66) bәt-mavvәdd-әn FUT-CAUSE.do-1MS

k̭at

that

+

ṱerә birds

+ ṱerә bәt-mavvәdd-әn-lun FUT-CAUSE.do-1MS-OBJ.3PL birds ‘I shall cause the birds to fly’

parx-i ~ fly-3PL k̭at

that

parx-i fly-3PL

našә (67) +sarbazә bәt-max-i soldiers FUT-hit-3PL people ‘The soliders will hit people’ (68) bәt-mavvәdd-әn FUT-CAUSE.do-1MS bәt-mavvәdd-әn-lun

k̭at

that

FUT-CAUSE.do-1MS-OBJ.3PL

+

sarbaz soldiers + sarbazә

әmax-i našә ~ hit-3PL people k̭at maxi našә

soldiers

that

hit-3PL people

524

Geoffrey Khan

‘I shall cause the soldiers to hit people’ There are no restrictions on the syntactic alignment of the caused embedded clause. So, the following types of clause, which cannot be made causative by a morphological causative, can be made causative by the periphrastic constructions: Ditransitive clauses: k̭at-ux (69) bәt-yavәl-lun FUT-give.3ms-obj.3pl to-2ms ‘He will give them to you’ k̭at-ux (70) bәt-mavvәddәn-lә k̭at yavәl-lun FUT-cause.do-3MS that give.3MS-OBJ.3PL to-2MS ‘I will make him give them to you’

Clauses with verbs taking a prepositional complement the referent of which is an affectee, e.g. biyy-ux (71) bәt-dak̭әr FUT-knock.3MS on-2MS ‘He will knock you’ (72) bәt-mavvәddәn-lә

k̭at

dak̭әr

biyy-ux

FUT-CAUSE.do-OBJ.3MS

that knock.3MS on-2MS ‘I shall cause him to knock you’

Semantic paramaters Dixon (2000) has drawn attention to the importance of documenting not only the syntactic alignment of causative constructions but also various semantic parameters of the verb and verbal arguments. When a language has more than one causative mechanism, these are generally distinguished by different semantic parameters. The semantic parameters relate to the verb, the causer and the causee. The parameters identified by Dixon (2000: 62) that have been found to be relevant for distinguishing the semantics of the various Neo-Aramaic causatives are the following: Verb:

(1) actionality (state/action) (2) transitivity

Causee

(3) control (4) volition (5) affectedness

Causative constructions in Neo-Aramaic (Christian Urmi dialect

Causer

525

(6) directness (7) intention

1. Actionality – Lexical causatives: The intransitive verb is always inchoative unaccusative expressing a change of state. The verb +k̭ṱl ‘to kill’ can be considered to be the lexical causative of the unaccusative verb myt ‘to die’. Lexical causatives in the form of the transitive alternant of labile verbs always have an inchoative unaccusative intransitive alternant. – Morphological causatives: Causatives formed by stem II are derived from unaccusative intransitives. Causatives formed by stem III, on the other hand, have no restriction as to actionality and may be derived from unaccusative verbs expressing a change of state and from agentive verbs expressing a dynamic action. –

Periphrastic causatives: There are no restrictions with regard to actionality.

2. Transitivity – Lexical causatives: The lexical causative +k̭ṱl is the causative of an intransitive verb (myt ‘to die’). The transitive alternant of a labile verb is by definition a causative of an intransitive. – Morphological causatives: Causatives from stem II can be applied only to intransitive verbs. Causative from stem III can be applied to intransitive and transitive verbs, but not to ditransitive verbs or verbs that express the affectee by means of a prepositional phrase. – Periphrastic causatives: These have no restrictions with regard to transitivity. 3. Control of the causee – Lexical causative: the causee of a lexical causative expressed by the direct object has no control. Likewise the causee object of the causative transitive alternant of labile verbs has no control. – Morphological causative. The same is the case with regard to the causee of morphological causatives of an intransitive verb expressed by stem II and stem III, in that the causee direct object does not have control. This applies both to the causative of non-agentive intransitives (e.g. plṱ II ‘to cause to go out’, m-cry

526

Geoffrey Khan

III ‘to cause to become short’) and agentive intransitives (e.g. m-brc III ‘to cause to kneel’). The causee of the stem III morphological causative of a transitive verb that is expressed by a prepositional adjunct, on the other hand, does have control, if the causee is agentive (e.g. bәt-mamxi-lun čalbә b-do naša ‘He will cause the dogs to be hit by that man’ (14)). Such agentive causees are typically human and introduced by the preposition b-. If the causee is an experiencer (e.g. bәtmaxzә ctava k̭a-babu ‘He will cause a book to be seen by his father’, i.e. ‘He will show a book to his father’ (16)), however, the causee is not in control. Prepositional b-phrases containing an inanimate noun in causative constructions such as + mәst-u b-nuyra (73) bәt-maxrәxx-әn-la FUT-CAUSE.singe-1MS-OBJ.3FS hair-3MS with-fire ‘I shall cause his hair to singe with fire’

do not express causees but are adverbial modifiers. – Periphrastic causatives: The causee of the verb m-∅vd must be animate and is in control, not only in a transitive clause but also in an intransitive one, e.g. (74) bәt-mavvәdd-i-lә FUT-CAUSE.do-3PL-OBJ.3MS ‘They will make him kneel’

k̭at

that

barәc kneel

4. Volition of the causee This parameter is closely related to that of control. Whenever the causee is in control according to the description in the last section, it can be assumed that this involves the causee acting willingly. This is reflected by the fact that the semantic range of the head verb m-∅vd in the periphrastic causative includes ‘persuade’. In several other NENA dialects, moreover, the verb ‘to allow’ is used in such constructions, e.g. C. Qaraqosh šwq ‘to let, to allow’: (75) bәd-šoq-әn banayә d-tor-i-lә FUT-leave.1MS builders COMP-break-3PL-OBJ.3MS ‘I shall make the builders break it.’ 5. Affectedness of causee It can be assumed that contexts in which the causee lacks control and volition, as described above, also involve a greater degree of affectedness of the causee. A significant semantic distinction can be identified, however, between the cau-

Causative constructions in Neo-Aramaic (Christian Urmi dialect

527

see direct object of a stem III lexical causative and the causee direct object of a morphological causative or stem II causative, in that the former is more affected than the latter. This is shown by the fact that if a stem III morphological causative has a reflexive pronoun as the causee the construction has the sense of ‘pretending’ rather than ‘causing’, e.g. m-myt III ‘to cause to die’ (76) bәt-mamyәt-la FUT-CAUSE.die.3MS-OBJ.3FS ‘He will pretend to die’ (77) m-šny ‘to cause to faint’ bәt-mašni-la FUT-CAUSE.faint.3MS-OBJ.3FS ‘He will pretend to faint’

ɟan-u selfFS-3MS

ɟan-u selfFS-3MS

This may be classified as a declarative type of causative (cf. Kulikov 2001: 892). Contrast this with a lexical causative with a reflexive causee, which express total affectedness of the causee: (78) bәt-+k̭aṱәl-la ɟan-u FUT-kill.3MS-OBJ.3FS selfFS-3MS ‘He will kill himself’ (lexical causative of myt ‘to die’) The object of morphological causatives of transitive verbs are, by contrast, fully affected, e.g. ɟan-u b-do (79) bәt-+mak̭ṱәl-la FUT-CAUSE.kill.3MS-OBJ.3FS selfFS-3MS by-OBJ.that ‘He will cause himself to be killed by that soldier’

+

sarbaz solidier

6. Does the causer act directly or indirectly? (i) Lexical causative: When the causee is the direct object of a lexical causative, the causer acts directly upon the object by manipulating it without mediation, e.g. (80) bәt-+k̭aṱәl-lә FUT-kill.3MS-OBJ.3MS ‘He will kill him’ [lexical causative of myt ‘to die’]) (ii) Morphological causative: Likewise, when the causee of an intransitive clause is made the direct object of a morphological causative, the causer acts directly, e.g.

528

Geoffrey Khan

(81) bәt-manpәl-lә FUT-CAUSE.fall.3MS-OBJ.3MS ‘He will cause him to fall’ The causer subject of a morphological causative of a transitive verb, on the other hand, does not act directly upon the object but through the mediation of the causee agent expressed by an adjunct prepositional phrase, e.g. (82) bәt-mamxi-lun b-do naša man fut-cause.hit.3MS-OBJ.3PL by-obl.that ‘He will cause them to be hit by that man’ (iii) Periphrastic causative: The causer in a periphrastic construction does not necessarily act directly by manipulating the object. This applies to the causative of both intransitive and transitive clauses, e.g. + palәṱ (83) bәt-mavvәddi-lә b-yavaltәt zuyzә k̭at FUT-CAUSE.do.3PL-OBJ.3MS by-giving money that leave.3MS ‘They will cause him to come out by giving him money’

7. Does the causer achieve the result accidentally or intentionally? (i) Lexical causative: The causer of a lexical causative does not necessarily act intentionally, since it may be inanimate, e.g. (84) poxa +tarranә bәt-patәx-lun FUT-CAUSE.open.3ms-OBJ.3PL wind doors ‘The wind will open the doors’ (ii) Morphological causative: Likewise, the causer in a morphological causative of an intransitive verb does not necessarily act intentionally, since it may be inanimate, e.g. ju-miyya (85) čipa bәt-+maṱbi-la in-water stone FUT-cause.sink.3MS-OBJ.3FS ‘The stone willl made her sink in the water’ (86) hak̭yat bәt-maɟxәcc-a-lun story FUT-cause.laugh-3FS-OBJ.3PL ‘The story will make them laugh’ The causer of a morphological causative of a transitive verb, on the other hand, is typically animate and so acts intentionally.

Causative constructions in Neo-Aramaic (Christian Urmi dialect

529

(iii) Periphrastic causative: The subject of the verb m-∅vd in the periphrastic construction is typically animate, so acts intentionally, whether the following clause is transitive or intransitive. In conclusion, we see that the syntactic and semantic properties of causative constructions in C. Urmi differ according to their degree of compactness, i.e. according to whether they are lexical, morphological or periphrastic causatives and, within the category of morphological causatives, according to their degree of morphological complexity (stem II being less complex than stem III).

References Alexiadou, Artemis, Elena Anagnostopoulou, and Florian Schäfer. 2006. “The Properties of Anticausatives Crosslinguistically.” In Phases of Interpretation, edited by Mara Frascarelli, 187–212. Berlin: Mouton. Comrie, Bernard 1981. Language Universals and Linguistic Typology: Syntax and Morphology. Oxford: Blackwell. Comrie, Bernard. 1976. “The Syntax of Causative Constructions: Cross-Language Similarities and Divergencies.” In The Grammar of Causative Constructions, edited by Masayoshi Shibatani, 261–312. Syntax and Semantics 6. New York: Academic Press. Dixon, Robert M. W. 2000. “A Typology of Causatives: Form, Syntax and Meaning.” In Changing Valency: Case Studies in Transitivity, edited by Robert M. W. Dixon and Alexandra Y. Aikhenvald, 30–83. Cambridge: Cambridge University Press. Dowty, David R. 1979. Word Meaning and Montague Grammar: The Semantics of Verbs and Times in Generative Semantics and in Montague’s PTQ. Synthese Language Library; Vol. 7. Dordrecht: D. Reidel. Goldenberg, Gideon. 2005. “Semitic Triradicalism and the Biradical Question.” In Semitic Studies in Honour of Edward Ullendorff, edited by Geoffrey Khan, 7–25. Leiden: Brill. ———. 2013. Semitic Languages. Features, Structures, Relations, Processes. Oxford: Oxford University Press. Göransson, Kathrin. 2014. “Causative-Inchoative Alternation in North-Eastern NeoAramaic.” In Neo-Aramaic and Its Linguistic Context, edited by Geoffrey Khan and Lidia Napiorkowska. Piscataway, NJ: Gorgias. Haspelmath, Martin. 1993. “More on the Typology of Inchoative / Causative Alternations.” In Causatives and Transitivity, edited by Bernard Comrie and Maria Polinsky, 87–120. Amsterdam: Benjamins. Ilani, Noga. 2013. “Inner Object.” In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan, Shmuel Bolozky, Steven E. Fassberg, Gary A. Rendsburg,

530

Geoffrey Khan Aaron D. Rubin, Ora R. Schwartzwald, and Tamar Zewi, vol. 2: 287–289. Boston: Brill.

Khan, Geoffrey. forthcoming. The Neo-Aramaic Dialect of the Assyrian Christians of Urmi. Kulikov, Leonid. 2001. “Causatives.” In Language Typology and Language Universals, edited by Martin Haspelmath, Ekkehard König, Wulf Oesterreicher, and Wolfgang Raible, 2:886–898. Berlin-New York: De Gruyter. Levin, Beth, and Malka Rappaport Hovav. 1995. Unaccusativity: At the Syntax-Lexical Semantics Interface. Cambridge, MA: MIT Press. Perlmutter, David. 1978. “Impersonal Passives and the Unaccusative Hypothesis.” Proceedings of the Annual Meeting of the Berkeley Linguistics Society 4: 157–189. Retsö, Jan. 1989. Diathesis in the Semitic Languages: A Comparative Morphological Study. Studies in Semitic Languages and Linguistics. Leiden: Brill. Saksena, Anuradha. 1980. “The Affected Agent.” Language 56 (4): 812–826. doi:10.2307/ 413490. Shibatani, Masayoshi. 1976. “Causativization.” In Syntax and Semantics, edited by Masayoshi. Shibatani, 5: 239–94. New York: Academic Press. Song, Jae Jung 2006. “Causative: Semantics.” Edited by Keith Brown. Encyclopedia of Language and Linguistics. Amsterdam London: Elsevier. Van Valin, Robert D. 2006. “Semantics in Role and Reference Grammar.” Edited by Keith Brown. Encyclopedia of Language and Linguistics. Amsterdam London: Elsevier.

Sonorant alternations in Muher Fekede Menuta, Hawassa University, and Ronny Meyer, Addis Ababa University

Abstract In this study, we deal with sonorant alternations, in particular between nn/ɲɲ~j and j~r, in Muher, an Ethiosemitic language of the Gurage cluster. Based on original fieldwork data on Muher supplemented by comparative data from other Ethiosemitic languages, we argue for a multi-causal origin of this phenomenon. From a synchronic point of view, the diachronic sound changes from geminated *ll to nn/ɲɲ, and singleton *l to j partly merged yielding the archiphonemes L (alternation nn~j) and Lʲ (alternation ɲɲ~j). In addition, the intense bilingualism with Ezha found among Muher speakers from Tattessa caused the contact-induced alternation j~r, as well as less frequent alternations including the sonorants l and n.

1

Introduction

The alternation between the sonorants r, n, l is a prominent phonological feature of most Gunnän Gurage varieties, except Kistane (cf. Hetzron 1977: 40–41). The alternation between r~n is commonly observed in verb inflection, in which the perfective base contains the nasal n (most frequently geminated) but the imperfective or jussive bases the trill r (always singleton), as in Mesqan bänna1 ‘he ate’ vs. jǝbära ‘he eats/will eat’, which are diachronically related to the Semitic root √blʕ ‘swallow’.2 In some Gurage varieties, this alternation is also found with nouns, as in Ezha neba ‘thief’ vs. jä-reba ‘for the thief’, which are related to Amharic leba ‘thief’. Generally, the lateral l is a very rare sound in Central and Peripheral Gurage varieties, in which it only occurs in a few lexical entities, or as result of a morphophonological processes, in which r comes in contact with another r or n at a morpheme juncture, as in Ezha /k’ar-no/ (thing-COP.3PL.M) > [k’allo] ‘they are things’ (cf. Hetzron 1977: 38).                                                                                                                 1

The vowel ä represents an open-mid central vowel in Ethiosemitic, probably close to IPA [ɜ], whereas ǝ refers to the mid central vowel.

2

The Gǝʿǝz data and etymologies for (Ethio)-Semitic roots are mainly taken from Leslau (1991), while the data for Gurage varieties other than Muher are from Leslau (1979a).

532

Fekede Menuta and Ronny Meyer

The sonorant alternation between r, n, and l is particularly well described for Chaha, where it is commonly analyzed as result of a diachronic neutralization of these sonorants into a single archiphoneme R (Degif 1996: 153; see also Hetzron 1977: 40). Rose (2007: 406–407), who is mainly concerned with the alternation of n~r in Chaha verbs, considers it the result of a diachronic process in which original geminated consonants were first “hardened” and then degeminated yielding pairs of allophones consisting of alternating strong vs. weak consonants. Regarding the sonorants, n is considered the strong counterpart of r, which diachronically represent geminated *ll or singleton *l, respectively. Obviously, this analysis does not account for the sonorant alternation in nouns, like the above-mentioned neba ‘thief’ vs. jä-reba ‘for the thief’, which also occurs in Chaha. Degif (1996; 2000) postulates synchronic morphophonological rules for the entire distribution of the sonorants in Chaha. According to him, the archiphoneme R has the allophones [n] in word-initial position or represents the result of nasalization (which occurs when an underlying geminated R is degeminated or in the coda of the penultimate syllable (Degif 1996: 154, 157)), while the allophone [r] occurs in almost all other environments. The allophone [l] represents geminated R at morpheme junctures. However, Degif (1996) also lists a number of exceptions which do not follow these rules. For most other Gurage varieties, only very general comparative data on sonorant alternations are available, like Hetzron (1977: 40–41, 49–52) or Leslau (1979a: xlvii–lv). In order to obtain a more detailed picture of sonorant alternations in Gurage varieties, this study describes their basic features in Muher – a Gurage variety mainly spoken in the Muher and Aklil District of the Gurage Zone. Isolated pockets of Muher speakers are also found in the Abeshge District near Welkite, the administrative capital of the Gurage Zone. The Muher data in this study represent the ädi-bet variety, which, in particular, is used in the Meqorqor area, but also spoken in Tattessa (for further details, see Meyer 2012). The Muher data discussed in this paper were gathered during several fieldwork stays in Tattessa and Welkite between 2001 and 2014. They are based on spoken texts and elicited verb paradigms provided by several native speakers (male and female, elders and youths).3                                                                                                                 3

We are very grateful to all the Muher speakers from Tattessa; in particular, we thank Abubakr Sherifo, our main consultant, for his patience and cooperation over all these years. Furthermore, we acknowledge the support of the Norhed project

 

Sonorant alternations in Muher

533

The phonemes r, l, n, ɲɲ, j and w represent the consonant sonorants in Muher. Except the labio-velar glide w, the phonological contrast between the remaining sonorants is blurred in several ways. Synchronically, two basic types of alternation are found. First, singleton sonorants, in particular j and r, and to a lesser extent also n or l, appear as free variants in a few verbs. As will be shown in Section 4, these alternations are due to recent language contact, mainly with Ezha and Amharic, and rather restricted to Tattessa Muher. Second, there are regular alternations between the geminated nasals nn or ɲɲ vs. non-geminated j in Muher in general. These alternations represented by the archiphonemes L for nn~j, and Lʲ for ɲɲ~j,4 as well as the diachronic sound change from *l to j are discussed in Section 3. The archiphonemes are most prominent in verb conjugation, where a consonant can be geminated in one verb form but singleton in another. Therefore, Section 2 will provide a concise overview about the formation of bases for verb inflection. Finally, Section 5 will summarize the main findings and forward our conclusions. 2

Verb inflection

The verb morphology in Muher – as in any other Semitic language – is nonconcatenative. Prototypically, a verb is conjugated by intersecting a consonantal root containing lexical information with a vocalic template encoding the inflectional bases for the perfective or imperfective aspect, or the jussive/imperative mood. The template also marks whether the second to the last radical in the root, i.e. C2, is geminated. The root contains information about the conjugation type of a verb, which also has an effect on the gemination of C2. A verb of conjugation type A is unmarked for gemination, but a verb in conjugation type B always geminates C2   (cf. Hetzron 1977: 70–71). Prototypically, a root contains three or four consonants. Due to diachronic sound changes, original consonants of a root can get lost or change into vocalic radicals (Hetzron 1977: 74–75), like the final a in the verb t’änna-m ‘be expensive’ in (1):5                                                                                                                                                                                                                                                                                                                                     Linguistic Capacity Building: Tools for the Inclusive Development in Ethiopia (2014– 2018) between Addis Ababa University, Hawassa University (Ethiopia), the University of Oslo, and the Norwegian University of Science and Technology. 4

Based on cognate items from related Ethiosemitic languages, we will show that these archiphonemes result from a sound change from a lateral *l to Muher j or nn/ɲɲ. Therefore, the archiphonemes are represented by the symbol L.

5

Following Prunet (1996), the vocalic radical a is represented by the capital letter A in the root (see also Podolsky 1991:26–27 for a discussion of other symbols).

534

Fekede Menuta and Ronny Meyer

(1) a. t’ännam

{t’änna}x-ä-m6

be_expensive.PFV-3SM-DCM ‘It is expensive.’

X = {√t’1n2A3 + EXPENSIVE

b. ant’äna X = {√t’1n2A3 + an-{t’äna}x-ä NEG-be_expensive.PFV.NEG-3SM EXPENSIVE ‘It is not expensive.’ c. wɔt’na wä-{t’na}x VN-be_expensive.JUSS ‘to be expensive’

X = {√t’1n2A3 + EXPENSIVE

C1äCC2äC3} PFV

C1äC2äC3} PFV.NEG

C1C2C3} JUSS

The Muher root √t’nA is most probably cognate with Gǝʿǝz √s’nʕ (with the variant √s’nʔ) ‘be strong’ (cf. Leslau 1979a: 622; 1991: 559). The change from Gǝʿǝz s’ to t’ is regular, as Muher lacks the ejective fricative s’ in its basic phoneme inventory. Furthermore, the glottal stop ʔ, as well as the pharyngeal fricatives ʕ and ħ in Gǝʿǝz roots regularly change into the vocalic radical A in Muher. The lexical information of the various verb forms of ‘be expensive’ in (1) is always encoded by the root √t’nA, while the templates provide grammatical information about the verb base. In (1a), for instance, the root √t’nA intersects with the template for the affirmative perfective C1äCC2äC3 yielding the base t’änna- to which an obligatory subject agreement index and other grammatical affixes can be attached. The second to the last root consonant in all three verb forms in (1) is the sonorant n, which occurs as geminated nn in (1a) but as singleton n in (1b–c) due to the gemination pattern associated with C2 in the templates. This alternation regularly occurs with all tri-consonantal verbs, like annäs- ‘be not enough’, t’ännäʔ- ‘be frightened’, zännäb- ‘rain’, zännäf- ‘hit’, but only with one other biconsonantal verb without a simplex stem, namely tɔ͡anna- ‘sit down’ (mediopassive stem) or ɔ͡anna- ‘put down’ (direct causative stem).

                                                                                                                6

The braces contain two intersected morphemes, namely the lexical root and the grammatical template. For a better understanding of the morphophonological processes, these two morphemes are given separately.

  3

Sonorant alternations in Muher

535

Sonorant alternation in archiphonemes

3.1 Archiphoneme L [nn~j] and the sound change *l > j Except two verbs, t’änna-m ‘be expensive’ in (1) and ɔ͡anna- ‘put down’ or its medio-passive stem, all recorded bi-consonantal verbs with the sonorant n as C2 exhibit the alternation geminated nn vs. non-geminated j, as shown in (2). In these verbs, C2  is the archiphoneme L representing nn~j, which stands in opposition to the phoneme /n/. Consequently, the verbs in (1) and (2) represent a minimal pair in which the alternation nn~n (for /n/) contrasts with nn~j (for /L/) as C2: (2) a. t’ännam {t’äLLa}x-ä-m hate.PFV-3SM-DCM ‘He hated.’ b. *ant’äja > ant’ɛ͡ä an-{t’äLa}x-ä NEG-hate.PFV.NEG-3SM

X = {√t’1L2A3 + C1äCC2äC3} HATE

PFV

X = {√t’1L2A3 + C1äC2äC3} HATE

PFV.NEG

‘He did not hate.’ c. wɔt’ja wä-{t’La}x VN-hate.JUSS

X = {√t’1L2A3 + C1C2C3} HATE

JUSS

‘to hate’ In (2b), the glide j as singleton representation of L merges with the preceding template vowel ä yielding ɛ, which, in turn, triggers the raising of the following vocalic radical A to ä resulting in the diphthong ɛ͡ä, i.e. *t’äLa>*t’äja>*t’ɛa>*t’ɛ͡ä. After consonants, however, the glide j is realized as [j] when followed by another vowel, as in (2c), or as vowel [i] when the glide occurs in the final syllable of the base, as in (3): (3)

wɔsk’i wä-{sk’j}x VN-hang.JUSS ‘to hang’

X = {√s1k’2L3 + C1C2C3} HANG

JUSS

536

Fekede Menuta and Ronny Meyer

The realization of the glide j after the vowel ä in (2b) may suggest that it actually could represent the floating feature ʲ. However, the lack of palatal coarticulation with the ejective velar k’ in wɔsk’i ‘to hang’ (instead of *wɔsc’) in (3) only allows an analysis as segment j. Other verbs with the archiphoneme L are given in List 1. All of them are biconsonantal verbs which end in the vocalic radical A: List 1 Verbs with the archiphoneme L (nn~j) Affirmative perfective Negative perfective an-C1äj2a C1änn2a-m (a) bänna-m am-bɛ͡ä (b) bʷɔnna-m am-bʷɛ͡ä (c) fänna-m am-fɛ͡ä7 (d) hänna-m ã-χɛ͡ä (e) männa-m am-mɛ͡ä

‘eat’ ‘go down, disappear’ ‘boil (itr.)’ ‘forbid’ ‘fill’

The cognate sets between Muher, Kistane and Gǝʿǝz in List 2 allude to the following diachronic sound changes. First, the pharyngeal fricatives ħ and ʕ in the Gǝʿǝz items lost their phonemic status and changed into the vocalic radical A, as now found in Kistane and Muher but also in most other Ethiosemitic languages. Beside the extinct Gǝʿǝz, they seem only to occur in the North Ethiosemitic languages Tigre and Tigrinya, as well as in Argobba. Second, geminated ll in Kistane changed to geminated nn in Muher. List 2: Cognate sets with the alternation ll~nn Muher Kistane (a) bänna-m bälla (b) fänna-m fälla (c) männa-m molla (d) hänna-m källa (e) t’änna-m t’älla

Gǝʿǝz bälʕa fälħa mälʕa kälʕa s’älʔa/s’älʕa

‘eat’ ‘boil (itr.)’ ‘be full’ ‘prevent, prohibit’ ‘hate’

Although the archiphoneme L is immediately followed by the vocalic radical A, the geminated nn in the Muher verbs in List 2 cannot directly originate from Gǝʿǝz lʕ or lħ. If the loss of the pharyngeal fricatives in Gǝʿǝz triggered gemina                                                                                                                 7

This verb has the variant amfäla, see List 16 and discussion in Section 4.

 

Sonorant alternations in Muher

537

tion of l, it would be geminated in all base forms of the verb, which is not the case, as shown in List 1. Furthermore, there are a few words with geminated nn in Muher which are related to geminated ll in cognates from other Ethiosemitic languages (but not always including Gǝʿǝz): List 3: Cognate sets with the mutation ll > nn word-medially Muher Other Ethiosemitic language (a) jannä

(c) ǝnnǝm

Amharic Ethiosemitic Amharic Zay Gafat

jallä √hlw8 jälläm ʔīllä9 (j)ǝllǝm10

(d) wɔnnäχʷä

Kistane

wälläho

(b) jännä

‘he/it that exists’ ‘he/it that does not exist’ ‘all, every’ ‘neighbor’

Thus, geminated nn in the Muher archiphoneme L and a few other instances is clearly connected with geminated ll in related languages. That means, neither Gǝʿǝz nor any other North Ethiosemitic language can be directly related to Muher as they do not geminate C2 in the perfective base but in the imperfective. In this case, the bases of the perfective and jussive – both lack gemination of C2 – would contrast with the imperfective base in which C2 is geminated. Example (2) shows that this is not the case; here rather geminated C2 of the perfective contrasts with singleton C2  of the imperfective and jussive. Only a few Muher entries were identified in which the word-initial or -final nasal nn corresponds to the lateral l in related languages:

                                                                                                                8 The existential verb √hlw in Ethiosemitic is a Type B verb, i.e. l as C2 is always geminated. 9 The negative existential verb probably represents a negated form of Ethiosemitic *hlw ‘exist’, as in Gǝʿǝz ʔi-hallo (Meyer 2005a: 147). 10 Cf. Leslau (1945: 48–49) in a transliteration which does not indicate gemination. Therefore, it remains unclear whether l in Gafat (j)ǝl(l)ǝm ‘all’ is geminated. In Teshome (1999: 64, 72, 77), the entry ‘all’ occurs in the phrase ba-jjämu ‘for all’ with geminated jj instead of l. This might indicate that in Leslau’s Gafat the original lateral was also geminated.

538

Fekede Menuta and Ronny Meyer

List 4: Cognate sets with the change l > n(n) word-initially or -finally Muher Gloss (a) nänn ‘upwards’

Other Ethiosemitic languages Gǝʿǝz *lʕl ‘be elevated’; Amharic *laj ‘on top’

(b) dänn ‘stomach, belly’ (c) ǝnnam ‘cow’

East Gurage däl (Leslau 1979b:590 f.) Gǝʿǝz lāhm (Leslau 1991:309); East Gurage lām; Kistane älam (Leslau 1979b:148 f.).

The condition of the change from l > n(n) in List 4 remains unclear. While original word-initial l changes to geminate nn in Muher ǝnnam ‘cow’, it appears as non-geminated n in nänn- ‘upwards’. Regarding dänn ‘stomach, belly’, it is unclear whether there was a diachronic stage in which the word-final *l was actually geminated. The glide j as part of the archiphoneme L results from the diachronic sound change singleton *l > j. List 5 contains cognates from Muher and Gǝʿǝz which exhibit this change in lexical roots. As mentioned in Section 2, when encircled by the central vowels ä or a, the glide j fuses with them to yield the diphthong ɛ͡ä. After consonants, however, it is realized as [j] or [i]. List 5 Cognate sets with alternating l and j in root morphemes Muher Gǝʿǝz ͡ (a) bäkk’ɛä-m /bäkk’äjä-m/ bäk’ʷälä ‘sprout’ (b) k’ätt’ɛ͡ä-m /k’ätt’äjä-m/ k’ätälä ‘kill’ ͡ (c) mässɛä-m /mässäjä-m/ mäsälä ‘seem’ (d) säkk’ɛ͡ä-m /säkk’äjä-m/ säk’älä ‘hang’ ͡ (e) bɛä-m /bajä-m/ bǝhlä ‘speak’ ͡ (f) hɛä-m /hajä-m/ kǝhlä ‘know (Gǝʿǝz: be able)’ (g) täsɛ͡ä-m /täsajä-m/ säʔälä ‘ask’ ͡ (h) wɛä-m /wajä-m/ wäʕälä ‘spend the day’ With regard to vocalized stems, only a few cognates were recognized which are given in List 6.

 

Sonorant alternations in Muher

539

List 6 Cognate sets with alternating l and j in vocalized stem morphemes Muher (a) k’ǝme

Amharic k’ǝmal

Gǝʿǝz k’(ʷ)ǝmal

(b) k’ǝt’e

k’ǝt’äl

k’(ʷ)äs’l

Muher: ‘leaf of ensete’; Gǝʿǝz/Amharic ‘leaf’

(c) gege

gǝlgäl

ʔǝgʷ(a)l

‘young of animals’

‘bug, louse’

Word-finally, the glide j in Muher fuses with the preceding central vowels *a and *ä to yield e. Thus, the glide j in Muher is a phoneme whose realization is highly depending on its immediate environment. 3.2 Archiphoneme Lʲ [ɲɲ~j] Most Muher verbs with the palatal nasal ɲɲ as C2 belong to the conjugation type B, in which C2 is always geminated. Examples for such verbs are given in List 7. List 7 Verbs with ɲɲ as invariable C2  in Muher (a) oɲɲä-m ‘cry’ (b) fuʔäɲɲä-m ‘whistle’ (c) tʃ’äɲɲä-m ‘give birth’ (d) tädwaɲɲä-m ‘be perplexed’ (e) zaɲɲä-m ‘dawn’ Among the verbs in List 7, only oɲɲä-m has a nominal form, ojat ‘cry (n)’, in which the glide j occurs instead of the geminated palatal nasal ɲɲ. This alternation between geminated ɲɲ vs. singleton j is represented by the archiphoneme Lʲ. There are only a few verbs in Muher that have a clear representation of the archiphoneme Lʲ as C2:

540

Fekede Menuta and Ronny Meyer

List 8 Verbs with the archiphoneme Lʲ Affirmative perfective a-C1äɲɲ2ä-m (a) a-fʷɔɲɲä-m

Negative perfective anna-C1j2ä anna-fujä

Verbal noun wa-C1i2 wa-fʷi ‘rest’

(b) a-gäɲɲä-m (c) a-mäɲɲä-m

anna-gjä –

wa-gi wa-mi

‘find’ ‘make’11

All verbs in List 8 lack a simplex stem; they are given in the direct causative stem that always starts with the vowel a-. Related to the verbs in List 8 is the verb k’ʷɔɲɲä-m ‘roast’ in which the palatal nasal has two realizations when non-geminated. It is [r] in most forms but also may appear as [j]: List 9 Alternation between [ɲɲ]~[r]~[j] (a) k’ʷɔɲɲäm /k’ʷäLLʲä-ä-m/ roast:PFV-3SM-DCM (b) jıʔɔru (c) jıʔojtt (d) jäʔuri

‘he roasted’

/j-k’ʷäLʲ-u/ 3SM-roast:IPFV-DCM

‘he roasts’

/j-k’ʷäLʲ-ʷ/ʲ-j-tt/ 3SG-roast:IPFV-IP-OJ.3SM-DCM /jä-k’ʷLʲ-ʷ/ʲ-j/

‘one roasts it’

3SG-roast:JUSS-IP-OJ.3SM

‘one may roast it’

In List 9 (c), the palatalization ʲ as part of the index for the impersonal -ʷ/ʲ may have triggered the change of r to j. However, this change does not occur in the impersonal in List 9 (d). Furthermore, just the reverse forms – jıʔɔritt instead jıʔɔjtt in (c), or jäʔɥ͡i instead jäʔuri in (d) – may sporadically occur. As the alternation ɲɲ~r was only recorded in this verb, it seems more plausible to consider it a lexicalized form which most probably entered Muher due to language contact with Ezha (see next section). Regarding the origin of the verbs in List 8, all of them lost their third rootconsonant. Obviously this consonant cannot be a pharyngeal or glottal consonant because it would be traceable as vocalic radical A. The only root consonants that may disappear diachronically without leaving a trace as vocalic radi                                                                                                                 11 This verb is defective in Muher; it only functions as supporting verb in phrasal verb constructions (cf. Meyer 2009).

 

Sonorant alternations in Muher

541

cal root-finally are the glides w or j. Furthermore, roots with final l must be considered because it may have changed into the glide j. Supposing these three consonants as underlying C3, six possible Gǝʿǝz cognates for the nine Muher verbs (cf. List 7, List 8, List 9) with the supposed archiphoneme Lʲ could be found in Leslau (1991): List 10 Possible Gǝʿǝz cognates for Muher verbs with the archiphoneme Lʲ Muher Gǝʿǝz (a) a-fʷɔɲɲä-m ‘rest’ fäläjä/fäläwä ‘separate’ (b) a-gäɲɲä-m ‘find’ (c) (d) (e) (f)

a-mäɲɲä-m oɲɲä-m tʃ’äɲɲä-m k’ʷɔɲɲä-m

‘make’ ‘cry’ ‘give birth’ ‘roast’

fälälä gäläjä (gänäjä mälälä wäjläjä/wäjläwä t’älälä k’äläwä

‘descend, prolapse’12 ‘reveal, disclose’ ‘bow down’) ‘plane, lengthen’ ‘mourn, lament’13 ‘be moist, be fertile, be fat’ ‘roast, parch’14

Most of the Gǝʿǝz cognates end in the glide j, which sometimes alternates with the glide w (see items (a), (d) and indirectly also (f) in List 10). Only in two, possibly three, items the last root consonant is l (items (c), (e) and (a)). The second consonant is the lateral l, with the possible exception of gänäjä in (b), as was suggested by Praetorius (1879:7) according to Leslau (1979a:286). Although it cannot be excluded that the glide j as lost root-final consonant triggered the palatalization of the preceding nasal n, such a process appears to be very unlikely given the other examples, in which ɲɲ results from *ll>nn and was palatalized through the lost root-final glide *j. If geminated ɲɲ represented an original n as C2 we would expect the non-geminated counterpart ɲ but not j.15 Further                                                                                                                 12 Leslau (1979a: 252) suggest a Cushitic origin for this verb, like Kambaata, Alaaba fayyeʔ or Sidaama fayyi ‘feel better, recover from illness’. 13 Could also be related to Zay awalähä or Silt’e awaläkä ‘speak, talk’ (Leslau 1979a: 652) with the initial sequence awa becoming o in Muher. 14 For Semitic languages outside Ethiopia, Leslau (1991:431) gives the root √k’lj, i.e. with final glide j instead of w. 15 The process of word-final denasalization is relatively unknown in Ethiosemitic. It is only reported for the northern Shewa variety of Amharic, in which word-final geminated ɲɲ may degeminate to ɲ and then even denasalize to j so that hed-ä-bbǝɲɲ (go.PFV-3SM-against-OJ.1SG) ‘he went away to my detriment’ as used in Addis

542

Fekede Menuta and Ronny Meyer

more, only the occurrence of the geminated palatal nasal ɲɲ is unpredictable in Muher, while non-geminated ɲ is always result of a synchronic palatalization process triggered by the non-segmental feature -ʲ. Thus, the archiphonemes Lʲ and L are most probably result of a similar diachronic development in Muher. As with L, original geminated ll changes to a geminated nasal, which, in the case of Lʲ, is palatalized to ɲɲ by the glide j as lost root-final consonant. 3.3 Conflicting cognates Contrary to the situation in Chaha as described in Degif (1996), the lateral *l did not complete merge with other sonorants but also functions as phoneme in Muher. Synchronically, there exist many lexical items in Muher which contain geminated ll, as shown in List 11: List 11: Cognate sets lacking the alternation *ll > nn

(e) gälläfä-m

Other Ethiosemitic languages Dobbi, Mesqan ella(-m) Kistane, Mesqan sälla(-m) Amharic alläbä Wolane baläʔä Silt’e bāläk’ä Mesqan bālläk’ä Kistane, Dobbi, Mesqan gälläfä(-m)

(f) sällämä-m

Amharic

(a) (b) (c) (d)

Muher ella-m sälla-m alläβä-m balläʔä-m

sällämä

‘intend, desire’16 ‘arrive’ ‘milk’ ‘be old, be wise’

‘be tall’ ‘become Muslim’

The items (c–f) in List 11 vis-à-vis List 1 and List 8 suggest that the change from underlying *ll to nn or ɲɲ occurs only in bi-consonantal verbs, i.e. it is always accompanied by the loss of the root-final consonant. If the root-final consonant is not lost, as is the case with the items (c–f) in List 11, the lateral ll as C2 is also retained and does not change to another sonorant. Furthermore, the sound change from *ll to nn or ɲɲ only occurs when the lost root-final consonant was the glide j (as shown in List 10), or the pharyngeal fricatives ʕ or ħ. The latter restriction can be deduced from the items (a–b) in                                                                                                                                                                                                                                                                                                                                     Ababa Amharic changes to hed-ä-bbǝ-ɲ or hed-ä-bb-i in northern Shewa (p.c. with Derib Ado, December 2014). 16 Leslau (1979a: 38) considers it a loanword from Cushitic, like Hadiyya heleʔā-kko.

 

Sonorant alternations in Muher

543

List 11, which retained the lateral l as C2. The Muher verb sälla-m ‘arrive’ seems to be related to Gǝʿǝz salaka ‘walk, go’ (cf. Leslau 1991:499), in which the rootfinal plosive k was first weakened to x (see Podolsky 1991: 29–32 for an overview on this phenomenon), and then to the vocalic radical A. Regarding ella-m ‘intend desire’, Leslau (1979a: 38) considers it a loanword from Cushitic, like Hadiyya heleʔā-kko, whose stem ends in the glottal stop. An argument against this assumption might be the Muher verb t’änna-m ‘hate’ in List 2 (e) whose Gǝʿǝz cognate ends in the glottal stop – but Leslau (1991: 554) also provides a variant ending in the pharyngeal fricative ʕ. Be that as it may, it seems that the various pharyngeal and glottal consonants of Gǝʿǝz which merged in the vocalic radical A in most current South Ethiosemitic languages do not belong to a totally uniform class of “weak consonants” as their diachronic loss triggered slightly different outcomes.17 The above-mentioned conditions for the sound change from *ll to nn do obviously not apply to the nominal items mentioned in List 3 (c–d) and List 4. Synchronically, verb roots as well as nominal stems in Muher also contain non-geminated l: List 12 Muher entries containing singleton l (a) (b) (c) (d) (e) (f)

fättälä-m χʷɔffälä-m t’ällälä-m ank’ulalla-m ank’ʷɔllälä-m ʃällälä-m

‘spin’ ‘be bored’ ‘be pure’ ‘stir while roasting’ ‘hover’ ‘promenade’

(g) k’ämäle (h) ʤale

‘guenon’ ‘males born at the same period’

(i) gamela

‘camel’

Generally, the change *l to j is found in verb roots in which *l occured as root-final consonant – either as last element (cf. List 6 and List 14) or followed                                                                                                                 17 A similar assumption has already been made for the Eastern Gurage variety Zay. Leslau (1951:215) observes that Zay verbs with Gǝʿǝz cognates ending in the glottal stop ʔ have a phonologically different form as compared to verbs whose Gǝʿǝz cognates have the glottal fricative h, the velar fricative x, or the pharyngeal fricatives ʕ or ħ as root-final consonant (cf. also Meyer 2005a:108).

544

Fekede Menuta and Ronny Meyer

by the vocalic radical A (cf. List 1). With regard to non-verbal entries, the situation is less clear. The three nouns in List 6 vis-à-vis the nouns (g–i) in List 12 may indicate that *l only changes to j stem-finally but not if followed by a vowel. There is at least one instance in which this sound change also occurred in another position, namely in Muher jidʤ ‘male child’, which is cognate with Amharic lǝdʤ ‘child’ (or wänd ‘male’), and Gǝʿǝz wäld ‘son, boy’. However, Muher jidʤ seems to be derived from an entity similar to Gǝʿǝz wäld rather than directly from Amharic jidʤ (cf. Meyer 2005b: 59). Labialization might be connected with the change *l to j, as the Gǝʿǝz cognate items in List 6 are optionally labialized whereas labialization is lacking with the nouns in List 12 (g–i). However, some of the verbs in List 12, which did not take part in the change from *l to j, also have a labialized consonant. Thus, it remains basically unclear why the sound change *l to j occurred in the verbs in List 5 but not with those in List 12. 3.4 Synopsis The sonorant alternations singleton j vs. geminated nn, or ɲɲ are result of two different diachronic sound changes from underlying geminated *ll to nn or ɲɲ, and singleton *l to j. With regard to verb roots, these two sound changes can be approximated as follows: (4) a. *ll> nn or ɲɲ only in C2 position, whereas i. *ll > nn if followed by the vocalic radical A as trace of lost *ʕ or *ħ ii. > ɲɲ if the glide *j diachronically occurred root finally iii. > ll elsewhere b. *l > j root finally, i.e. in C2 position if not followed by another consonant or in C3 position The sound change from singleton *l to j is usually found root finally (cf. List 5) in Muher but also occurs as part of the archiphonemes L in which singleton *l is followed by the vocalic radical A. The sound change *ll to nn is found in roots whose last consonant was originally a pharyngeal fricative (ʕ or ħ) which lost its consonantal status but remained as vocalic radical A part of the root (see List 2). When the glide j – either as root consonant on its own, or as result of the palatalization of *l – lost its status as consonant and got deleted, it left a trace through the feature palatalize-

 

Sonorant alternations in Muher

545

tion resulting in the geminated palatal nasal ɲɲ (see List 10). In all other instances, geminated *ll remains as such in Muher. Except the archiphoneme Lʲ, all other occurrences of the palatal nasal are phonologically conditioned and predictable. As a result, the palatal nasal ɲɲ is the only phoneme in Muher that is always geminated. The two sound changes (a) and (b) in (4) became interconnected when they occur with root consonants that can be geminated or not depending on the word form yielding the archiphonemes L and Lʲ. 4

Contact-induced alternation between sonorants

Almost all tri- and quadri-consonantal verbs with r as final root consonant have a free variant with the final glide j instead of r in the Muher variety spoken in Tattessa. This alternation seems to be infrequent or lacking in other Muher varieties (cf. Leslau 1981; 1979a: liv f.). The verbs with the root-final sonorant r in the first column in List 13 are the common Muher forms, while the verbs in the second column, in which r was replaced by j, are particular to the Muher variety spoken in Tattessa. List 13: Free alternation between r~j in Tattessa Muher ‘cause to lose money (in business)’ (a) akässärä-m or akässɛ͡ä-m18 (b) k’att’ärä-m or k’att’ɛ͡ä-m ‘knot’ (c) täβattärä-m or täβattɛ͡ä-m ‘be separated’ (d) tʃ’әrä-m or tʃ’ıjä-m ‘stink (v.)’ (e) wɔttärä-m or wɔttɛ͡ä-m ‘stretch, tighten’ When confronted with the free alternation r~j in these verbs, most Muher speakers from Tattessa prefer the forms in r. In free speech, however, the change r > j frequently occurs in an almost unpredictable manner. With the affirmative perfective verbs in List 13, the alternation r~j is very common, but r is still favored over j. However, the jussive and imperative forms of these verbs strongly prefer j over r. With a few verbs, just the reverse pattern is found. The perfective simplex gәdäddɛ͡ä-m ‘become big (several plants)’, for instance, is almost never uttered as gәdäddärä-m with r as final root consonant, but predominantly appears as jus                                                                                                                

18 The diphthong ɛ͡ä is the phonetic realization of the sequence /äjä/, see Section 3.1.

546

Fekede Menuta and Ronny Meyer

sive simplex jä-gdädәr ‘may it become big’. The perfective simplex ʃäggɛ͡ä-m ‘change’, rarely uttered as ʃäggärä-m, may be realized in the imperative optionally as sägi-nni or sägǝr-ǝnni [change.JUSS.2SM-BEN.OJ.1SG) ‘Change (2SM) (it) for me!’ In addition, the verb k’ʷɔɲɲä-m ‘roast’ (cf. Section 3.2, especially List 9) prefers j over r in most conjugated forms. The preference for j over r is also found in certain derived stems. The indirect causative and medio-passive stems of wɔttärä-m~wɔttɛ͡ä-m ‘stretch, tighten’ (List 13 (e)), for instance, most frequently contain the glide j, i.e. at-wɔttɛ͡ä-m ‘make tight’ (indirect causative stem) or tä-wɔttɛ͡ä-m ‘be tight’ (medio-passive stem). While the simplex nätt’ärä-m ‘be purified’ often has the sonorant r as root-final consonant, its causative derivations a-nätt’ɛ͡ä-m ‘purify’ (direct causative) and at-nätt’ɛ͡äm ‘order to purify’ (indirect causative) usually end in the glide j. The alternation r~j can also be observed in verb forms vis-à-vis derived nominals. Usually, the nominal derivation invariably occurs with root-final r. The verb gәdäddɛ͡ä-m ‘become big (several plants)’, for instance, has the derived adjective gǝdǝddǝr ‘big (PL)’ with root final r. The abstract noun related to the verb hättɛ͡ä-m ‘make a dam, hinder to cross’ is only uttered as hätärat ‘off-time (in quarrel)’. The synchronic alternation r~j is peculiar to the Muher variety spoken in Tattessa, and here in particular in the verbal domain. This is a remarkable fact as the use of either r or j as root-final consonant is semantically distinctive in a few verbs across all Muher varieties: List 14: Minimal pairs distinguished by r vs. j (a) nätt’ärä-m (√nt’r) ‘purify’ nätt’ɛ͡ä-m (√nt’l > √nt’j) ‘separate (a mother from her child)’ (b) k’äbbärä-m (√k’br) k’äbbɛ͡ä-m (√k’bl > √k’bj) (c) kässärä-m (√ksr) kässɛ͡ä-m (√ksl > √ksj)

‘plant, bury’ ‘lack, be not full’ ‘loose money (in business)’ ‘(i) try hard, make an effort; (ii) become charcoal’

 

Sonorant alternations in Muher

547

As indicated in List 14, original *l underwent a sound change to j (cf. also Section 3.1) while original *r is generally retained as r in Muher.19 In addition, an optional sound alternation r~j can be observed in the Muher variety spoken in Tattessa so that the diachronic sound change *l > j is merging with the free alternation r > j in this variety. An important factor for the free alternation r > j in Tattessa Muher is probably the intense multilingualism of its speakers with Ezha,20 in which the original singleton sonorants *l, *n and *r all merged into r root-finally (cf. Polotsky 1938: 142–143). Tattessa-Muher speakers may easily observe that root-final r (diachronically related to *l) in many Ezha verbs is realized as j in Muher. Consequently, they may perceive the alternation r~j as a main distinctive feature between Ezha and Muher, and extend it to all cognate verbs ending with r in order to keep the two varieties formally apart. This kind of hypercorrection seems to be the reason for the alternation n~r~j in the Tattessa Muher verb sǝnässɛ͡ä-m~sǝrässärä-m ‘fine down, plane’. The two variants are related to Ezha sǝrässärä-m, Mesqan/Dobbi sǝlässälä, or Wolane sǝnässälä (Leslau 1979a: 544). Although the relation between l, n, r and j can clearly be noticed, the chronology of the sound changes remains unclear. The Muher realization sǝnässɛ͡äm could have its immediate origin in Wolane sǝnässälä, i.e. the nasal n was not affected but Wolane l changes to Muher j. Wolane sǝnässälä or Mesqan/Dobbi sǝlässälä would yield Ezha sǝrässärä-m, which could have been incorporated as such into Muher. Furthermore, the root-final alternation n~j was once recorded in the verb athäddänä-m~athäddɛ͡ä-m ‘order to cover a roof’ whereby the cause for the alternation is unclear. A similar unexpected alternation in which root-final j was re                                                                                                                 19 In Central and Peripheral Gurage varieties, like Ezha or Chaha, original *l was changed into r so that these verbs became homonymous with the verb roots ending in r, i.e. the pairs of morphologically distinct Muher verbs from (a) to (c) in List 14 have only the form ending with r for each pair in Ezha and Chaha. On the other hand, some Muher nominals end with a stem-final palatal element while the sonorant r occurs in cognate items in other Gurage varieties (cf. Leslau 1979a:lv), like Muher fʷe ‘on top’, ʤe ‘small jar’, ʃɛ͡ɛt ‘spider’ vs. fʷär (Chaha, Ezha), ʤär (Ezha), or ʃäret (Dobbi), respectively. 20 Today, Ezha is the dominant language in Tattessa, while Amharic is commonly used in nearby Welkite and at school. In Tattessa, Muher, if at all, is mainly used in the family domain so that the active knowledge of this language is declining.

548

Fekede Menuta and Ronny Meyer

placed by n occurred in the imperative of the verb t’äkk’ɛ͡ä-m ‘hide’ when followed by object indexes, as shown in (5).21 (5) a. t’әʔәnu {t’k’j~n}x-ʲ-u hide.JUSS-2SF-OJ.3SM ‘Hide (2SF) him!’ b. t’әʔәnǝnn {t’k’j~n}x-nn hide.JUSS.2SM-OJ.3SM ‘Hide (2SM) him!’

X = {√t’1k’2j3 + C1C2C3} HIDE

JUSS

X = {√t’1k’2j3 + C1C2C3} HIDE

JUSS

The Ezha forms for these verbs are t’äkk’ärä-m ‘hide’ (but Mesqan, Kistane t’äkk’änä), and häddärä-m ‘cover a roof’ (Mesqan häddänä, Kistane käddänä). Another Ezha influence on the r~j alternation in Tattessa Muher is most probably found in verb inflection. The 2SF subject index with imperfective and imperative verbs contains the non-segmental feature ʲ, which is suffixed to the verb base. In Ezha, but basically not in Muher, this feature triggers the palatalization of root-final r to j. In Tattessa Muher, too, there is a strong tendency among many speakers to palatalize root-final r with 2SF subjects, as in Ezha, while other speakers do not: List 15: Optional palatalization of root-final r (1) a. k’ǝβǝr /k’br/ plant.JUSS.2SM b. k’әβi or c’ǝβǝr22 /k’br-ʲ/ plant.JUSS-2SF c. c’әbru /k’br-ʲ-u/ plant.JUSS-2SF-OJ.3SM

‘Plant (2SM)!’ ‘Plant (2SF)!’ ‘Plant (2SF) it!’

                                                                                                                21 In postvocalic position, the glottal stop ʔ is the regular allophone of the singleton ejective k’. 22 The consonant [c’] is the phonetic realization of the palatalized velar ejective k’ʲ.

 

Sonorant alternations in Muher

(2) a. t’ǝra

/t’rA/ call.JUSS.2SM

b. t’ijä or t’ǝrä c. t’әro

/t’rA-ʲ/ call.JUSS-2SF /t’rA-ʲ-u/ call.JUSS-2SF-OJ.3SM

549

‘Call (2SM)!’ ‘Call (2SF)!’ ‘Call (2SF) him!’

The optional palatalization of root-final r to j is only found in absolute wordfinal position in the (b) examples in List 15. When a further suffix is attached, as in the (c) examples, the root-final sonorant is uttered as r. However, in the formation of impersonal verbs, which contains the feature -ʷ/ʲ and an additional object index (see List 9 (c–d)), palatalization of r to j is very frequent. In three verbs with geminated nn as C2 in the affirmative perfective, the sonorants r, l, and j alternate in an unpredictable way as singleton C2: List 16 Alternation between nn~r/l/j Affirmative perfective Negative perfective an-C1är/l/j2a C1änn2a-m (a) fänna-m aɱ-fäla or aɱ-fɛ͡ä (b) anna-m ann-ära (c) fʷɔnna-m aɱ-fʷɔra or aɱ-fʷɔla

‘boil (itr.)’ ‘defecate’ ‘have sexual intercourse’

The verbs under (a) and (b) in List 16 are loanwords, which are only rarely used in Muher discourse. The verb fänna-m in (a) referring to boiled coffee, tea, milk, etc. is borrowed from Amharic fälla ‘boil (itr.)’; the native Muher verb is wɔtt’am ‘(a) go out; (b) be ready after boiling’. In the perfective, the Amharic fälla ‘boil (itr.)’ shows the regular change ll>nn for C2. As singleton, however, C2   either follows the regular Muher pattern by changing to j or keeps the non-geminate l of Amharic. The verb (b) in List 16 seems also to represent a loanword from Amharic, which entered Muher via Ezha. The sonorant r as C2 in Amharic arra ‘defecate’ regularly yields geminated nn or singleton r in Ezha. If this verb were directly borrowed from Amharic to Muher, this sound change would not have occurred. The verb (c) in List 16, fʷɔnna-m ‘have sexual intercourse’, is a taboo word that may be disguised by treating it optionally like a loanword from Ezha or Amharic by realizing singleton n as either r or l.

550

5

Fekede Menuta and Ronny Meyer

Summary and conclusion

The sonorant alternations in Tattessa Muher are result of two distinct language change processes, namely the diachronic sound changes *ll>nn or ɲɲ and *l>j, as well as the alternation between r~j and to a lesser extent between l~r, l~j and n~j induced by extensive language contact with Ezha and Amharic. The language internal diachronic sound changes did not affect all occurrences of the lateral *l(l) but were restricted to certain phonological environments (see Section 3.4). Consequently, the lateral l did not completely merge with the remaining sonorants ɲɲ, n, r, and j, nor completely fused these sonorants with each other. All sonorants retained their phonemic status in Muher. Usually, diachronic sound changes do not yield the loss of phonological contrast between segments synchronically. However, due to the non-concatenative morphology, which prevails in major parts of Muher verb formation and noun derivation, a specific consonant in a root can be either singleton or geminated in individual word forms. If this consonant happens to be related to the lateral *l(l), the two different diachronic sound changes intermingle resulting in the archiphonemes L and Lʲ which account for the synchronic alternation between the geminated nasals nn or ɲɲ with singleton j. Language contact is the other process responsible for sonorant alternations in Tattessa Muher, as discussed in Section 4. The r~j alternation is clearly due to bilingualism with Ezha. Here two distinct phenomena were observed: First, in Ezha, a sound change from underlying *l to r in verb roots occurred instead of the sound change from *l to j in Muher. Consequently, Ezha lost the contrast between original r and *l>r yielding homonymous verbs, as those in List 14, which are still morphologically distinguished through the contrast r vs. j in Muher. Tattessa-Muher speakers extended this peculiar phonological difference as a kind of hypercorrection to other occurrences of r as root-final consonant in Muher resulting in the r~j alternation. Second, some Tattessa Muher speakers extended the palatalization of r to j, i.e. a morphophonological rule from Ezha, to their Muher speech. This is most commonly found in bare imperatives with second person female subjects, but also in the impersonal form of a verb. Hypercorrection might also be the reason for the rare alternation n~j. In other cases, the glide j in original Muher verbs may alternate with the trill r in the same verb when it is directly borrowed or code-switched from Ezha. This, particularly, applies to verbs in which geminated nn alternates with singleton r in List 16, while verbs with the alternation geminated nn vs. singleton l are

 

Sonorant alternations in Muher

551

partly adapted from Amharic. The resulting very rare alternation between j~r~l occurs due to the overlap of borrowings or code-switches from these various languages. Borrowing is most probably also the cause for the reintroduction of the lateral l in a position where it actually should have changed into j, i.e. with the verbs in List 12, as well as with the irregular entities with the sound change *ll to nn mentioned in List 3c–d and List 4. To conclude, the sonorant alternations in Muher have various causes that cannot be explained through a strict synchronic or diachronic approach but only by a combination of the two. Furthermore, language contact, as one cause of variation, need not be restricted to the current situation but might have occurred in various periods of the history of Muher with changing contact languages. In a broader perspective, the diachronic sound change involving the lateral *l(l) as well as language contact are also of relevance for the sonorant alternations in other Gunnän Gurage varieties. It seems that the diachronic fusion of the lateral *l with another sonorant and the resulting loss of phonological contrast between l, r and n is most advanced in the Chaha group, but almost absent in Dobbi and Mesqan, and completely lacking in Kistane. In the Chaha group, the alternation is not restricted to the gemination/non-gemination of the same consonant in different word forms as found in Muher, but also regularly occurs in other morphophonological contexts. Furthermore, the distinction between alveolar nasal nn vs. palatal nasal ɲɲ does not occur in most of the varieties within the Chaha group; here the two nasals merged into nn. Although the sonorant alternation in general is a prominent phonological feature for most Gunnän Gurage varieties, its actual alternation patterns and conditions are not well-known for most of the varieties. Consequently, their trigger and the sequence of their advancement from one variety to another are still enigmatic. Even other Ethiosemitic languages may show sporadic traces of a diachronic sonorant alternation *l(l) > n in the lexicon (e.g. Amharic wänd ‘male’ is related to Gǝʿǝz wäld ‘son, child’ (cf. Leslau 1991: 613)) or in grammar (e.g. the benefactive suffix on verbs is -n in Eastern Gurage but -ll in Amharic (cf. Meyer 2005a: 102)). Future research should investigate whether the abundance and prominence of sonorant alternations in Gunnän Gurage is due to a common ancestor (which is strongly rejected in Hetzron 1972), or mutual cultural and linguistic contacts in a compact convergence area.

552

Fekede Menuta and Ronny Meyer

Abbreviations 1, 2, 3

First, second, third person

BEN

Beneficiary

DCM

Declarative affirmative clause

IP

Impersonal

IPFV

Imperfective base

JUSS

Jussive base

NEG

Negation

OJ

Object index

PFV

Perfective base

SF

Singular-feminine

SG

Singular

SM

Singular-masculine

VN

Verbal noun

References Degif Petros Banksira. 1996. “Sonorant alternations in Chaha.“ In Grover Hudson (ed.). Essays on Gurage language and culture, dedicated to Wolf Leslau on the occasion of his 90th birthday, 153–173. Wiesbaden: Harrassowitz. Degif Petros Banksira. 2000. Sound mutations: The morphophonology of Chaha. Philadelphia / Amsterdam: John Benjamins. Hetzron, Robert. 1972. Ethiopian Semitic: Studies in classification. Manchester: Manchester University Press. Hetzron, Robert. 1977. The Gunnän-Gurage languages. Napoli: Istituto Orientale di Napoli. Leslau, Wolf. 1945. Gafat documents: Records of a South-Ethiopic language: Grammar, text and comparative vocabulary. New Haven, Connecticut: American Oriental Society. Leslau, Wolf. 1951. “Archaic features in South Ethiopic.“ Journal of the American Oriental Society 71(4): 212–230. Leslau, Wolf. 1979a. Etymological dictionary of Gurage (Ethiopic). Vol. III: Etymological section. Wiesbaden: Harrassowitz. Leslau, Wolf. 1979b. Etymological dictionary of Gurage (Ethiopic). Vol. II: English-Gurage index. Wiesbaden: Harrassowitz. Leslau, Wolf. 1981. Ethiopians speak: Studies in cultural background part IV – Muher. Wiesbaden: Harrassowitz. Leslau, Wolf. 1991. Comparative dictionary of Gǝʿǝz: Gǝʿǝz-English / English-Gǝʿǝz with an index of the Semitic roots. Wiesbaden: Harrassowitz. Meyer, Ronny. 2005a. Das Zay: Deskriptive Grammatik einer Ostguragesprache (Äthiosemitisch). Köln: Köppe.

 

Sonorant alternations in Muher

553

Meyer, Ronny. 2005b. “The morpheme yä- in Muher.“ Lissan: Journal of African Languages and Linguistics 19(1): 40–63. Meyer, Ronny. 2009. “The quotative verb in Ethiosemitic languages and Oromo.“ In Joachim Crass and Ronny Meyer (eds.). Language contact and language change in Ethiopia, 17–42. Köln: Köppe. Meyer, Ronny. 2012. “The use of Muher demonstratives for references in space and discourse.“ Afrika und Übersee 91: 161–202. Podolsky, Baruch. 1991. Historical phonetics of Amharic. Tel-Aviv: Baruch Podolsky. Polotsky, Hans Jürgen. 1938. “Étude de grammaire gouragué. “ Bulletin de la Société de linguistique de Paris 39: 137–175. Praetorius, Franz. 1879. Die amharische Sprache. Halle: Verlag der Buchhandlung des Waisenhauses [Reprint: Hildesheim: Olms, 1970]. Prunet, Jean-François. 1996. “Guttural vowels.“ In Grover Hudson (ed.). Essays on Gurage language and culture, dedicated to Wolf Leslau on the occasion of his 90th birthday, 175–203. Wiesbaden: Harrassowitz. Rose, Sharon. 2007. “Chaha (Gurage) morphology.“ In Alan S. Kaye (ed.). Morphologies of Asia and Africa. Volume 1, 403–427. Winona Lake, Indiana: Eisenbrauns. Teshome Yehualashet. 1999. “Preliminary phonological description of Gafat.“ Zena Lissan 1(1): 62–79.

Arabic or Ṣūrayt/Ṭūrōyo Sina Tezel, Uppsala University

Abstract The present article deals with seven illustrative case studies in order to demonstrate which criteria one has to follow in studying a category of Arabic borrowings in Ṣūrayt/ Ṭūrōyo. Arabic borrowings derived from Arabic cognate roots in a Semitic context are of two categories, namely those with phonological correspondences and those without such correspondences. In the former case, the phonological shape of a root is the most important criterion to take into consideration, while in the latter case the emphasis of analysis is on the form, the meaning, and the bǝġaḏkǝfaṯ-issue. The article deals with the latter group of cognates, showing that non-assimilated borrowings are relatively easy to identify, while the assimilated borrowings of this kind offer difficulties on several levels, especially when an Aramaic/Syriac and Arabic cognate root has the same meaning.

Some introductory words Studying the Arabic borrowings in Ṣūrayt/Ṭūrōyo one encounters problems of varying degrees. One can divide the Arabic loanwords in two main categories. There are Arabic loanwords having no Aramaic/Syriac cognates in a Semitic context, and there are Arabic loanwords having Aramaic/Syriac cognates in a Semitic context. The latter can, in turn, be divided in two categories, namely those with systematic phonological correspondences and those having no such correspondences. In studying the loanwords with systematic correspondences the phonological shape of a root is the most important criterion to follow, and in a language such as Ṣūrayt/Ṭūrōyo, with a relatively conservative consonantal system, the phonological shape of a root with systematic phonological correspondences is decisive for the conclusions.1 In studying the other category of cognates one has to take into account a range of other criteria. In such cases one has to look at the form, the meaning, and in certain cases also at the bǝġaḏkǝfaṯissue. In the case of non-assimilated loanwords, derived from such cognate roots, such criteria can provide significant help (cf., for example, Ṣūrayt/Ṭūrōyo maqbāra sub the case qbr/qwr), but in the case of assimilated loanwords of this 1

For borrowings with phonological correspondences, see S. Tezel 2011.

Arabic or Ṣūrayt/Ṭūrōyo

555

category sometimes all the criteria mentioned are not sufficient to draw firm conclusions. For example, how do we proceed in our discussion when we can see that Ṣūrayt/Ṭūrōyo has qawro, “a grave”, and qabro (especially of Jesusʼ grave) and nowadays also qabro in Ṣūrayt/Ṭūrōyo songs. Is the form qabro due to influence from Arabic or are there other underlying reasons for this doublet? The matter becomes more complicated when we deal with verbal roots of this category. For example, which approach shall we take in a case such as Ṣūrayt/ Ṭūrōyo ʿbr, “to enter, come in”, when an investigation of the Arabic dialects in the region shows that Anatolian Arabic ʿbr also has the same meaning. Further, how shall we explain the fact that NENA for the same root has ʾwr, and Mlaḥso ʿvr, that is, the b in the old root ʿbr has been subjected to fricativization due to the bǝġaḏkǝfaṯ-shift in these languages but not in Ṣūrayt/Ṭūrōyo. In the following pages an attempt will be made to answer these and similar methodological issues by demonstrating seven case studies. Discussion of seven illustrative case studies 1. ḥbk To the interesting Ṣūrayt/Ṭūrōyo verbal roots belongs ḥbk, “to darn, mend (of socks and clothes)”, underlying ḥbәk- in ḥbәkle (verbal stem I). It is attested in Ritterʼs Grammatik in the following instance: ḥbәkla ag-gurwe “sie stopfte die strümpfe”.2 According to the information I have obtained, it is a common word. In Syriac ḥbk/ḥḇḵ in its pәʿal form has meanings such as “to mix, mingle; to join, unite; to confuse, confound”, which Brockelmann etymologically connects with Arabic ḥbk,3 which in its faʿala form means “to weave well and tight; to twist, twine, tighten”.4 Among the principle Semitic languages only Syriac and Arabic seem to have such a root. The Syriac ḥbk/ḥḇḵ is assumed to have a cognate in Syriac ḥbq, “to surround”.5 An interesting different etymological connection of the Syriac ḥbk/ḥḇḵ with Arabic ʿbk is found in Thesaurus, referring to Bernstein.6 Arabic ʿbk in the phrase ʿabakahu means “he mixed it (a thing with a thing)”, according to Lane.7 Turning to Ṣūrayt/Ṭūrōyo root ḥbk and comparing it to the 2

Ritter 1990: 72.

3

Brockelmann 1928: 210a.

4

Wehr 1976: 154b.

5

Sokoloff 2009: 409b; Brockelmann 1928: 210a.

6

Thesaurus Syriacus 1981: col. 1174.

7

Lane 1863-93: 1941a.

Sina Tezel

556

Syriac root ḥbk/ḥḇḵ, we are facing two problems on the semantical and phonological level. As has been indicated, the meaning “to darn, mend” of the Ṣūrayt/ Ṭūrōyo ḥbk is not indicated for the Syriac root ḥbk/ḥḇḵ in the Syriac lexicons at hand. Besides, if a native case, the Ṣūrayt/Ṭūrōyo ḥbk considerably diverges from the bәġaḏkәfaṯ-rule, according to which one had to except a root form *ḥwx and a verbal stem *ḥwәx- in *ḥwәxle. The inflectional base of the Ṣūrayt/Ṭūrōyo verbal stem ḥbәk- is namely a passive participle in status absolutus of the pattern qәṭīl, that is, if a native word it should reflect Syriac ḥәḇīḵ, which in Ṣūrayt/ Ṭūrōyo would have the form *ḥwәx or *ḥwīx. That the half vowel ә, preceding the ḇ in the Syriac ḥәḇīḵ, is lacking in the Ṣūrayt/Ṭūrōyo ḥbәk- would not be considered a problem, as the loss of the vowel in question in native verbal stems of this pattern is a rule rather than an exception; cf. Ṣūrayt/Ṭūrōyo ḥwәṭin ḥwәṭle “to beat; to dash, throw against the ground”, the form ḥwәṭ- reflecting the old ḥәḇīṭ. Looking at the NENA-lexicon there we can note a root-form xwx, indicated in Maclean with the meanings “to mix; to join, to introduce into; to darken, obscure, confuse”, considered it a reflex of the old ḥbk/ḥḇḵ.8 In this case, the connection with the Syriac ḥbk/ḥḇḵ is apparent both on semantical and phonological level. This is especially true of the phonological shape of the root xwx, in which the old b and k have originally been subjected to fricativization in accordance with the bәġaḏkәfaṯ-rule. It has already been suggested that the bәġaḏkәfaṯ-issue in Ṣūrayt/Ṭūrōyo is a very important criterion in determining whether a Ṣūrayt/Ṭūrōyo word is native or foreign. Despite its importance, this alone can not underlie the conclusions. The meanings of a word are at least equally important. Comparing the meaning of the Ṣūrayt/Ṭūrōyo ḥbk to the meanings of the Arabic ḥbk we can immediately see that the meaning of the former is closest to those of the latter. This gives rise to the question whether the Ṣūrayt/Ṭūrōyo ḥbk is not a borrowing from the Arabic one. The answer much depends on the technical meaning “to darn, mend”, which could not be noted for the Arabic ḥbk either. In literary Arabic, the technical meaning “to darn, mend” seems to be signified by rfʾ. Thus rafaʾa ṯ-ṯawba means “he repaired”, or “mended”, or “darned, the garment, where it was rent”; or “he closed up what was rent in his garment, drawing the parts together”, by “texture [with the needle”, i.e. “darning]”, according to Lane.9 In dealing with Arabic borrowings into Ṣūrayt/Ṭūrōyo, Arabic dialects in the region are more 8

Maclean 1901: 91a.

9

Lane 1863-93: 1117b.

Arabic or Ṣūrayt/Ṭūrōyo

557

important than literary Arabic. An investigation of the sources at hand of some Arabic dialects in the region shows that the Iraqi Arabic uses the normal word for “to sew”, xyṭ,10 also to mean “to darn”, and the sense “to darn, mend” in Levantine Arabic is signified by rty.11 No ḥbk could be noted for Anatolian Arabic in the sources at hand, the Kinderib dialect having rfy, “mit einem Faden umnähen”;12 cf. the above mentioned literary Arabic rfʾ. To sum up this case, the semantical and phonological criteria of the Ṣūrayt/ Ṭūrōyo ḥbk are in favour of a borrowing of Arabic ḥbk, though the latter in our sources of Arabic does not provide the technical meaning “to darn, mend”. 2. npq/nfq Aramaic/ Syriac has a root npq/nfq, which Brockelmann connects with nfq in Arabic and Ethiopic.13 The same root occurs also in Ugaritic, according to Leslau, but in Gordon Ugaritic npq is indicated with a question mark.14 In its pәʿal form, Syriac npq/nfq has predominant meanings such as “to go out, issue” or “proceed forth; to come up, come out; to pass; to turn out”.15 In Arabic nfq in its faʿala and faʿila forms has meanings such as “to sell well, find a ready market; to be brisk, active; to be used up, be spent, run out”, be exhausted; to die, perish”, according to Wehr.16 The native npq/nfq has been retained in Ṣūrayt/Ṭūrōyo both as npq and nfq (henceforth nfq, which is more common),17 where it underlies two verbal stems, namely nāfәq with a predominant meaning “to go out”, and mawfaq- in mawfaqle or mōfaqle18 as its causative form. Besides the word mәnāfәq, all other words with the radicals npq/nfq with the notion “to go out”, attested in my sources are 10 See Clarity et al. 2003: 50a, having the following example: “did you darn my socks”? xayyaṭṭī-li ǧwārībi? 11 Barthélemy 1935-1969: 270; Frayḥa 1973: 62b; Stowasser and Moukhtar 2004: 60a. The last source has the following example: “did you darn my socks?”, retētī-li žrābāti? 12 Jastrow 2005: 59a. 13 Brockelmann 1928: 438b. 14 Leslau 1987: 388b; Gordon 1965: 446a. No Ugaritic npq could be noted in del Olmo Lete and Sanmartínʼs Dictionary. 15 Payne Smith 1903: 345a-b. 16 Wehr 1976: 987b. 17 For its occurrence both as npq and nfq, see Ritter 1990: 631ff. 18 For the assimilation of the first radical n, see A. Tezel 2003: 170-171.

558

Sina Tezel

derivates of the native root npq/nfq and are thus genuine. We can not approach the genuineness of this root by applying the bǝġaḏkǝfaṯ-issue, as the old p nowadays is mostly pronounced as f in Ṣūrayt/Ṭūrōyo, regardless its position. The most reliable criteria to follow in this case are the meanings of the root which, more or less, agree with those of the Syriac npq/nfq. A Ṣūrayt/Ṭūrōyo word with somewhat different root consonants, namely mawqo, “a tumour”,19 in which the f became w, also seems to be a derivative of the native root npq/nfq; cf. Western Syriac nafqō, “a tumour”.20 Turning to mәnāfәq, which means “hypocrite”, it is attested once in Ritter in the phrase: bīšo mәnāfәq d-kīt mīnux, “schlechter heuchler, der du bist”.21 The form and meaning of this word does not point to a native one but rather to a borrowing from Arabic munāfiq, “hypocrite, dissembler”,22 which in form is an active participle of the verbal stem III, nāfaqa, “to dissemble, play the hypocrite”. That the meanings of this verbal stem and its derivatives considerably differ from those quoted under Arabic nfq above is explained as a result of a borrowing. Both munāfiq and the verbal stem nāfaqa with the meaning quoted are namely considered borrowings from Geʿez.23 The plural form munāfiqūn of munāfiq is a common word in the Qurʾān, where it, both as masc. and fem., occurs thirty-three times, according to Jeffery.24 As a Qurʾanic word it was also borrowed into other languages, showing its spread in the region. 3. ʿbr The root for “to enter, come in”, in Ṣūrayt/Ṭūrōyo is termed ʿbr, which underlies two verbal stems and their derivatives, namely ʿābәr (stem I) and maʿbar- in mabʿarle, “to make” or “let enter, bring in”.25 The Aramaic/Syriac proper root for “to enter”, ʿll, is not attested in any verbal form in Ṣūrayt/Ṭūrōyo but in some nouns. The other Neo-Aramaic languages show a mixed picture. NENA, Western Neo-Aramaic and Mlaḥso have reflexes of both the old ʿll and ʿbr, with

19 For its occurrence, see also Ritter 1979: 324. 20 For details, see A. Tezel 2003: 170- 71. 21 Ritter 1979: 335. 22 Wehr 1976: 988a. 23 Nöldeke 1910: 48; Jeffery 1938: 272; Leslau 1987: 388b. 24 Jeffery 1938: 272. 25 Ritter 1990: 154.

Arabic or Ṣūrayt/Ṭūrōyo

559

the meaning “to enter”, which in Neo-Mandaic is rather signified by dyš.26 Lingering with Ṣūrayt/Ṭūrōyo ʿbr and its cognates in the other Neo-Aramaic languages we can state that the reflex of the old ʿbr/ʿḇr in both Ṣūrayt/Ṭūrōyo and Western Neo-Aramaic is ʿbr, in NENA ʾwr and in Mlaḥso ʿvr, all with the meaning “to enter, come in”.27 In the case of NENA ʾwr, it also means “to pass, pass over, pass by, cross, go over; to surpass; to pass through; to leave; to turn away from; to pass away, to be finished; to transgress, to break (a commandment); to come upon, to take hold of”, according to Maclean.28 Also in the case of Ṣūrayt/Ṭūrōyo, the sense “to pass” comes to the surface in a phrase: ʿābәr ʾūtānūro, “(the heating time of) the oven has passed”. Ṣūrayt/Ṭūrōyo ʿbr has also a sense “to break into” in the phrase: ʿābīri ʿal-rīšayye, “they broke into their house”, its literal meaning being “they entered on their head”. Except the meaning “to enter, come in”, all other meanings indicated do not offer problems as compared to the meanings of Syriac ʿbr/ʿḇr, “to pass on, by, over” or “beyond; to go to (any place), to invade; to transgress, violate; to pass away, vanish; to surpass”, and the like.29 The meaning “to enter, come in” of this root in the NeoAramaic languages is fraught with a problem. As has been suggested, the meaning in question is not indicated for the Syriac ʿbr/ʿḇr. The same seems to be true of ʿbr in the other early attested Aramaic languages. To be sure, it is not difficult to imagine a semantic development “to pass, cross over” > “to enter, come in”, or rather a semantic development “to go to” > “to enter”, but the question is why such a meaning is lacking in the early attested Aramaic languages. Besides, the meaning “to enter, come in” is indicated for at least some Anatolian Arabic dialects; cf. the following meanings of the Anatolian Arabic ʿbr in Vocke and Waldner: “hintreten, eintreten, eindringen, hineintreten, hineinstürzen, hereinkommen (durch), hinübergehen, vorbeigehen, überholen; vergehen (Tag)”.30 In literary Arabic, the root ʿbr in its basic stem has predominant meanings such as “to cross, to pass over, to ferry; to pass away”. The meaning “to enter, come in” could not be found for ʿbr in the consulted sources of 26 Mutzafi 2014: 181. Neo-Mandaic ʾll, “to go”, was previously assumed to be a derivative of ʿll, “to enter”, but new analyses rather state a derivative of the root ʾzl (see Mutzafi 2014: 32, referring to Morgenstern). 27 For Western Neo-Aramic ʿbr, see Bergsträsser 1921: 1-2; for NENA ʾwr, see Maclean 1901: 234; For Mlaḥso ʿvr, see Jastrow 1994: 155. 28 Maclean 1901: 234. 29 Payne Smith 1903: 398b. 30 Vocke and Waldner 1982: 273.

560

Sina Tezel

literary Arabic. The same is true of ʿbr in the consulted sources of the Arabic dialects in the region, except Anatolian Arabic. It thus seems that the meaning “to enter, come in” in Arabic is restricted to some dialects of Anatolian Arabic. This gives rise to the question whether a borrowing is involved. That the NENA has the root form ʾwr and Mlaḥso has the root form ʿvr clearly shows that the old b has been subjected to the bәġaḏkәfaṯ-shift in these languages. This is a strong indication of their being genuine cases, as foreign roots borrowed into these languages and Ṣūrayt/Ṭūrōyo normally are not the subject of the bәġaḏkәfaṯ-shift unless they are inherited via Aramaic/Syriac. Lingering with the bәġaḏkәfaṯ-rule, which is a very important criterion in determining whether a root is native or foreign, and considering the phonological shape of Ṣūrayt/ Ṭūrōyo ʿbr we can see that the b in this case has not been subjected to the bәġaḏkәfaṯ-shift unlike the cases in NENA and Mlaḥso. That being so, one wonders whether the Ṣūrayt/Ṭūrōyo ʿbr is genuine or not. Much of the answer lies in the inflectional base of the verbal stem of the Ṣūrayt/Ṭūrōyo ʿābәr, which morphologically reflects an adjective or a participle adjective in status absolutus of the pattern qaṭṭīl. Thus, provided that the Ṣūrayt/Ṭūrōyo ʿābәr is native, it actually reflects an adjective such as Syriac ʿabbīr, “passing over; past”.31 Degemination in Ṣūrayt/Ṭūrōyo is a rule rather than an exception. A degeminated Ṣūrayt/ Ṭūrōyo form of the old ʿabbīr would exactly yield ʿābәr as follows: ʿabbīr > *ʿābīr > ʿābәr, with compensatory lengthening of a > ā and with the change ī > ә in the closed syllable. Thus, the occurrence of the b in ʿābәr in intervocalic position does actually not contradict the bәġaḏkәfaṯ-rule, as it was originally geminated. As a rule, degemination does not lead to new changes related to the bәġaḏkәfaṯissue. However, some of the verbal stems of the same pattern show that they have undergone a secondary fricativization. Some examples: ʿātәq, “it became old” (from *ʿattīq), lābәṭ “it budged” (from *labbīṭ), but sāwәʿ “he was full (of food)” (from *sabbīʿ), rāwәʿ, “it lied down” (from *rabbīʿ). Thus, the verbal stems of the pattern in question do not show a homogenous picture and therefore the application of the bәġaḏkәfaṯ-issue in this case is not beyond all doubt. As has already been indicated, even Western Neo-Aramaic has ʿbr (also: ʿpr), but this is the normal pronunciation of the old b in this language. The occurrence of the root ʿbr with the meaning “to enter, come in” in Western Neo-Aramaic could help us to answer whether the meaning in question has originated in Aramaic/ Syriac or in Arabic. Provided that ʿbr in Levantine Arabic and the other major 31 Sokoloff 2009: 1064.

Arabic or Ṣūrayt/Ṭūrōyo

561

Arabic dialects in the region have not the meaning in question, a borrowing from Anatolian Arabic into Western Neo-Aramaic is improbable. This, and the fact that the root forms ʿbr, ʾwr (< ʿwr) and ʿvr in all Neo-Aramaic languages have the meaning concerned is suggestive of its being originated in Aramaic/ Syriac. The semantic development seems to be “to go to” > “to enter, come in”; cf. the meanings of bwʾ, “to come, to go, to enter”.32 If that is the case, this means that the meaning “to enter” of the Anatolian Arabic ʿbr is a borrowing from Aramaic/Syriac and Ṣūrayt/Ṭūrōyo ʿbr is genuine. 4. qbr/qwr A root qbr with a basic meaning “to bury, intern” occurs in all the principal Semitic languages.33 It has been retained in Ṣūrayt/Ṭūrōyo in the form qwr, where it underlies qwәr- in qwәrle, with the meaning “to bury”.34 The transition b > ḇ > w is ultimately due to the bәġaḏkәfaṯ-shift and is quite in accordance with the bәġaḏkәfaṯ-rule, as the verbal stem qwәr- reflects a historical form such as Syriac qәḇīr, the passive participle in absolute state. In the Ṣūrayt/Ṭūrōyo stem qwәr-, the short vowel ә, originally preceding the w, is thus dropped. All derivatives with the radicals qwr, with the notion “to bury”, are derived from this root and are thus genuine. We can decide the genuineness of this verbal root by applying the bәġaḏkәfaṯ-issue, as borrowings in Ṣūrayt/ Ṭūrōyo are, as a rule, not the subject of the bәġaḏkәfaṯ-shift. Genuine is also the word for “a grave, tomb”, qawro, which occurs in the language as a doublet, the other form being qabro. The latter is used to denote “the tomb/grave” (of Jesus, saints and Church Fathers), while qawro is the normal word for “a grave, tomb”. The form qabro, which today mostly occurs among the older generation, is attested twice in the following instances in Ritter: …ʾū-navī, dū-xūri, hūle-li šamʿo b-īḏi, u mbēle-li sū-qabro d-māfiryōno Šamʿin, “…der enkel des bischofs eine kerze in der hand und führte mich zu dem grabe des erzbischofs Schamʿun”; …u ʾazzīno, nšiqli ʾū-qabro, ʾāḏīʿat?, “-und ich ging und küßte das grab- du verstehst?”.35

32 Hoftijzer and Jongeling 1995: 146. 33 See, among others, Brockelmann 1928: 644a; von Soden 1965-1981: 912b; Leslau 1987: 440a; Koehler and Baumgartner 2001: 1064a. 34 For the occurrence of qwr and its use, see Ritter 1990: 449. 35 Ritter 1967: 380, verses 21 and 30, respectively; Ritter 1979: 389.

Sina Tezel

562

As we can see, the form qawro is genuine and the occurrence of w in postvocalic position is in accordance with the bәġaḏkәfaṯ-rule, while b in qabro contradicts the rule in question. As the word for “a grave, tomb” in Arabic is termed qabr, one wonders whether the Ṣūrayt/Ṭūrōyo qabro has been influenced from the Arabic form. Such a question can not be answered with certainty. However, the indications are those the irregularity in this case is not due to a borrowing from Arabic qabr but rather due to that we are dealing with a high church term and to the fact that the West Syrians today no longer apply the bәġaḏkәfaṯ-rule to an etymological b and p consistently. The former is, as a rule, pronounced as b and the latter as f.36 This is also evident from the use of qabro in newly composed Ṣūrayt/Ṭūrōyo profane songs; cf. the use of qabro in the phrase: warde nәfqi ʿalqabri, “flowers are springing up on my grave”,37 in which qabro   has been introduced from Western Syriac. The word for “cemetery, graveyard”, maqbara/maqbāra,38 on the other hand, is not genuine. It is a feminine substantive ending in -a, which, with the exception of a few cases, occurs in foreign nouns. Its form and meaning are rather suggestive of Arabic maqbara(t), with the same meaning.39 Foreign words are, as a rule, not subjected to the bәġaḏkәfaṯ-shift. The word for “cemetery” in Syriac is bēṯ qәḇūrē, which is not attested in Ṣūrayt/Ṭūrōyo as a genuine case, and, if it occurs nowadays, it has been introduced from Syriac into the language; cf. NENA bēṯ qwūrē, bēṯ qawrāwāṯē, baṯ quwrā, “graveyard”, in Maclean.40 5. qym/qwm An Aramaic/Syriac root qwm/qym, with predominant meanings “to stand, stand up; to rise”, is common Semitic.41 From the root form qym, with the meanings quoted, Ṣūrayt/Ṭūrōyo has two verbal stems, namely qāyәm and māqәm, the latter being the causative form. A third verbal stem mqāyam- in mqāyamle “to evaluate” is used mostly by Ṣūrayt/Ṭūrōyo-speakers familiar with Arabic. A fourth verbal form in the language is termed mqāwәm (passive/reflexive), “to 36 A. Tezel 2003: 21, n. 7. 37 Thus in a profane song Izla Shafirto by Aboud Zazi. 38 Ritter 1979: 315. 39 See also Ritter 1979: 315, considering the same word Arabic. 40 Maclean 1901: 31b. 41 In Akkadian, qâmum is considered Canaanite, von Soden 1965-1981: 896b.

Arabic or Ṣūrayt/Ṭūrōyo

563

happen”. It is also used in the phrase mqāwamme ʿal ḥḏōḏe, “they started to fight with each other”.42 Turning to the verbal stems qāyәm and māqәm, there are no reasons to call in question their genuineness. The same can be said of two derivatives of the native root qwm/qym, namely qyamto, “resurrection”, and qәmṯo, “stature”, which, judging from their respective forms and meanings, closely reflects Western Syriac qәyōmtō and qawmәṯō, respectively. Also from the bәġaḏkәfaṯ point of view, they are regular. To be sure, the ṯ in qәmṯo on the surface contradicts the bәġaḏkәfaṯ-rule, as it is in a post-consonantal position, but as has been demonstrated by the Western Syriac qawmәṯō it was originally preceded by the half-vowel ә, which is now dropped in qәmṯo. There remain the verbal stem mqāyam- and the verbal form mqawam-, the genuineness of the both may be called in question. Already Ritter considers the verbal form mqāwam Arabic. For the sense “to happen” of mqāwәm, Ritter also refers to Kurdish qäwimîn, which Chyet in his dictionary rather compares to Neo-Aramaic qavam and mqawo:me,43 which occurs in NENA; cf. (m)qăwim, “to happen, occur”, in Maclean, comparing it to the root qym.44 In Sabar, qwm, “be possible, occur; have sudden bad weather”, is considered Arabic, with a comparison to the root qym.45 It seems that the Arabic root qwm in its qāwama-stem once was spread in the region. In the case of the sense “to fight”, a borrowing from the Arabic stem qāwama is apparent, while the origin of the sense “to happen”, which is attested in three languages, namely Ṣūrayt/Ṭūrōyo, NENA and Kurdish, is problematic, for it could not be noted in the Arabic sources at hand. In any event, as for the Ṣūrayt/Ṭūrōyo case, two important signs, namely its meanings and phonological shape qwm, show that is a borrowing in the language. A genuine case should consistently have the form *mqāyam- and thus not mqāwam, for the old II-w/y verbal roots seem to have survived in the language only as II-y unless they are denominatives.46 If we now turn to mqāyam “to evaluate”, morphologically and phonologically this stem is not fraught with problems. Even semantically one could relate it to Syriac qayyem, which, among other things, means “to price, value”.47 How42 Ritter 1990: 476. 43 Chyet 2003: 486a. 44 Maclean 1901: 194. 45 Sabar 2002: 275. 46 A. Tezel 2003: 37. 47 Payne Smith 1903: 495a.

564

Sina Tezel

ever, there are some obstacles in the way for considering it genuine. As has been mentioned, this verbal form is mainly used by Ṣūrayt/Ṭūrōyo-speakers familiar with Arabic. Judging from its absence in Ritterʼs Grammatik, Ṣūrayt/ Ṭūrōyo-speakers from Ṭūr ʿAbdīn was probably not familiar with the word in that area. Besides, the technical sense “to evaluate” in itself is a semantic innovation even in Arabic qayyama, which do not even exist in Wehrʼs Dictionary, having rather qawwama, but with, among others, the senses “to estimate, appraise, value”.48 Thus, the indications are in favour of a borrowing even in this case. If it is introduced from Syriac and used by the purists, then the word is to be considered a neologism in Ṣūrayt/Ṭūrōyo. Ultimately of Arabic origin are also all non-assimilated nouns derived from the root qwm/qym. Here belong, for example, qīma/qime “worth, value”; mәqām, “musical tune, key; manner”; taqwīm,   “calendar”.49   The most important criteria followed in such cases   are the absence of the nominal endings -o, -to and -ṯo which do not occur in non-assimilated foreign nouns. 6. qrb/qrw A root qrb with a basic meaning “to be near, draw near” occurs in all the principal Semitic languages.50 It has been retained in Ṣūrayt/Ṭūrōyo in the form qrw, underlying two verbal stems, namely qāru (< *qārәw < *qarrīw < *qarrīḇ), with a predominant meaning “to come near”, and mqāraw- in mqārawle, “to come near; be just round the corner”, cf. u-saṯwo mqāraule, ”der winter steht vor der tür”; but also “to celebrate the liturgy, mass”; cf. mqāraule, “er las die messe”, in Ritter.51 In the present, kō-qōraw (with the preposition l, “to”) it also means “to be a relative/ to be related to”; cf. u-zʿūrāno k-qurāuloxyo?, “ist dieser junge mit dir verwandt?”, in Ritter.52 The transition b > ḇ > w in both stems is in accordance with the bәġaḏkәfaṯshift, as the w in the stems in question occurs in postvocalic position. All deri-

48 Wehr 1976: 799a. 49 For qīma/qime see Ritter 1979: 411 and for mәqām, see Ritter 1979: 315. 50 See, among others, Brockelmann 1928: 691b; Koehler and Baumgartner 2001: 1972ab; von Soden 1965-1981: 915b. 51 Ritter 1990: 284-285. 52 Ritter 1990: 284.

Arabic or Ṣūrayt/Ṭūrōyo

565

vatives with the radicals qrw, with the notions “to be near; to be a relative”, are derivatives of the same root and are thus genuine. An interesting doublet, derived from the root form qrb, where the radical b is not changed to w, is qurbōno, and qurban/qurbān.53 The former denotes “the Holy Communion; an offering”, and the latter is mostly used as a polite word, with meanings “my dear!, please!”,54 but also “sacrifice, offering”; cf. its use in the following instance in Ritter: ḥōlo qurbân dīḏux!, “Onkel! Ich will dein opfer sein!”.55 The form qurbōno, which closely reflects Western Syriac qūrbōnō, is genuine, judging from its forms and meanings, while the form of qurban/qurbān is rather suggestive of Arabic qurbān, “sacrifice, offering, immolation, oblation; Mass (Chr.),56 which is considered a borrowing from Aramaic/Syriac by several scholars.57 As has been suggested, the form qurban/qurbān is not used as an ecclesiastical term in Ṣūrayt/Ṭūrōyo but rather as a polite phrase; cf. qûrbân and qûrbânâ in Maclean; however without considering qûrbân Arabic.58 The preservation of the b in the form qurbōno, in which -ōno does not belong to the root, is in accordance with the bәġaḏkәfaṯ-rule, as the b occurs in post consonantal position. 7. rʿy The root rʿy with a basic meaning “to graze, feed” occurs in all the principal Semitic languages.59 It has been retained in Ṣūrayt/Ṭūrōyo just as rʿy, where it 53 It is also attested in the form qurba, in which the n is dropped, see Ritter 1979: 425, saying “ar/k”, that is Arabic/Kurdish. 54 Ritter 1979: 425. 55 Ritter 1969: 374-375, verse 282. 56 Wehr 1976: 755a. 57 See, among others, Nöldeke 1910:37, Leslau 1987: 440b (quotes Nöldeke); Jeffery 1938: 234-235. Jeffery writes “The Muslim authorities take the word as genuine Arabic, a form fuʿlān from qrb to draw near...Undoubtedly it is derived from a root qrb to draw near, approach, but in the sense of “oblation”, it is an Aramaic development, and borrowed thence into the other languages. In O. Aram. we find qrbn in this sense, and the Targumic qrbn’, Syr. qūrbānā: are of very common use...”. In any case, as an ecclesiastical term in Arabic it is a borrowing from Syriac. 58 Maclean 1901: 274b. 59 Brockelmann 1928: 737b; Koehler and Baumgartner, 2001: 1258ffb; Leslau 1987: 459b; von Soden 1965-1981: 976b.

566

Sina Tezel

underlies two verbal stems, namely rʿē- in rʿēle (stem I),60 with the same meaning, and marʿē- in marʿēle as its causative form (stem III).61 Two substantives, derived from the same root, namely ruʿyo, “shepherd”, and marʿīṯo, “diocese”, are native, judging from their forms. They closely reflect Western Syriac rōʿyō and marʿīṯō, respectively. The change ō > u in the closed syllable of the Ṣūrayt/Ṭūrōyo ruʿyo is a normal one. A third substantive, namely marʿa, “grassland, pasture place”,62 does not seem to be genuine. It is a feminine substantive ending in -a, which normally occurs in foreign nouns. Its form and meaning are rather suggestive of dialectal Arabic marʿa (literary Arabic: marʿa(n)), with the same meaning. In this case, the form of the word quoted is decisive and thus not its meaning, for in Syriac, too, we can note some substantives, derived from the same root, with the same or a similar meaning. Thus, Costaz indicates rʿīṯā, reʿyā (alone or with bēṯ, thus bēṯ reʿyā) and marʿīṯā (also with bēṯ, thus: bēṯ marʿīṯā), with, among other things, the same meaning as that of Arabic marʿa(n), that is “pasture”.63 There are no traces of the Syriac words just quoted with the same meaning in Ṣūrayt/Ṭūrōyo, having, as said, marʿīṯo but only with the meaning “diocese”. Speaking of the form of the word, in Maclean we find marîyâ (Syriac characters mrʿyʾ), with the same meaning as that of Ṣūrayt/Ṭūrōyo marʿa. Maclean relates this marîyâ to Arabic marʿāt,64 which in Arabic is a synonym of marʿa(n).65 In the case of marîyâ in Maclean, there are reasons for being suspicious of its not being borrowing from Arabic. The form of the word rather points to a native formation, on which already Nöldeke writes: “Für das ebenso gebildete marʿīṯā Herde, Gemeinde wird in der Bedeutung Weide die Form mariʿyā (marrîâ) gebraucht, welche im Grunde mit jener identisch ist”.66 Turning to Western Syriac marʿīṯō, we note marʿīṯ (sic!) among the Arabic words given to Syriac marʿīṯā in Thesaurus.67 This may indicate that the Syriac 60 See Ritter 1990: 333, where rʿēle, “er hat geweidet”, and p. 336, where rʿalle u-gēlo, “sie weideten das gras ab”, occur. 61 Ritter 1990: 405. 62 For its occurrence, see Ritter 1979: 317. 63 Costaz 1963: 349b. 64 Maclean 1901: 200a. 65 Bustānī 1870: 341c. 66 Nöldeke 1868: 105b. 67 Thesaurus Syriacus 1981: col. 3948.

Arabic or Ṣūrayt/Ṭūrōyo

567

word once was borrowed into Christian Arabic just as marʿīṯ, used as a religious technical term to denote “diocese”.

References Barthélemy, Adrien. 1935-1969. Dictionnaire arabe-français; Dialects de Syrie: Alep, Damas, Liban, Jérusalem. Paris: Geuthner. Bergsträsser, Gotthelf. 1921. Glossar des neuaramäischen Dialekts von Maʿlūla. Leipzig: in Kommission bei F. A. Brockhaus. Brockelmann, Carl. 1928. Lexicon Syriacum. Editio secunda aucta et emendata. Halis Saxonum (2. unveränderter reprographischer Nachdruck der 2. Auflage, Halle an der Saale). Tübingen: Niemayer. Bustānī, Buṭrus, 1870 (reprint: Beirut 1983). Muḥīṭ al-muḥīṭ, qāmūs muṭawwal li-l-luġa alʿarabiyya. Beirut: Maktabat Lubnān. Chyet, Michael L. 2003. Kurdish-English Dictionary. New Haven, CT: Yale University Press. Clarity Beverly E. et al. 2003. A Dictionary of Iraqi Arabic: English-Arabic. Washington, DC: Georgetown University Press. Costaz, Louis. 1963. Dictionnaire syriaque-français. Syriac-English Dictionary. Qāmūs siryānī ʿarabī. Beirut: Imprimerie Catholique. Frayḥa, Anīs. 1973. Muʿǧam al-ʾalfāẓ al-ʿāmmiya – English title: A Dictionary of Non-Classical Vocables in the Spoken Arabic of Lebanon (in Arabic). Beirut: Librairie du Liban. Gordon, Cyrus H. 1965. Ugaritic Textbook. Rome: Pontifical Biblical Institute. Hoftijzer, Jan and Karel Jongeling. 1995. Dictionary of the North-West Semitic Inscriptions. 2 vols. Handbuch der Orientalistik, part 1: Nahe und der Mittlere Osten, 21.1-2. Leiden/New York/Cologne: E. J. Brill. Jastrow, Otto. 1994. Der neuaramäische Dialekt von Mlaḥsô. Wiesbaden: Harrassowitz. ––– 2005. Glossar zu Kinderib. Wiesbaden: Harrassowitz. Jeffery, Arthur. 1938. The Foreign Vocabulary of the Qurʾān. Baroda: Oriental Institute. Koehler, Ludwig and Walter Baumgartner. 2001. The Hebrew and Aramaic Lexicon of the Old Testament: Study Edition, 2 vols. Leiden: Brill. Lane, Edward W. 2003. Arabic-English Lexicon, 8 vols., London 1863–1893. Reprinted by the Islamic Texts Society (UK), 2 vols. London: Wiliams and Norgate. Leslau, Wolf. 1987. Comparative Dictionary of Geʿez (Classical Ethiopic): Geʿez-English/ English-Geʿez. Wiesbaden: Harrassowitz. MacLean. Arthur J. 1901. A Dictionary of the Dialects of Vernacular Syriac as Spoken by the Eastern Syrians of Kurdistan, North-West Persia, and the Plain of Mosul. With Illustrations from the Dialects of the Jews of Zakhu and Azerbaijan, and of the Western Syrians of Tur ‘Abdin and Ma‘lula. Oxford: Clarendon Press. Mutzafi, Hezi. 2014. Comparative Lexical Studies in Neo-Mandaic. Leiden: Brill.

568

Sina Tezel

Nöldeke, Theodor. 1868. Grammatik der neusyrischen Sprache am Urmia-See und in Kurdistan. Leipzig: T. O. Weigel. ––– 1910. Neue Beiträge zur Semitischen Sprachwissenschaft. Strassburg: Trübner. Payne Smith, J. 1903. A Compendious Syriac Dictionary. Oxford: Clarendon Press. Payne Smith, Robert (ed.). Thesaurus Syriacus, I–II, Oxford 1879-1901 (Nachdruck der Ausgabe Oxford 1879–1901, Hildesheim, New York: G. Olms 1981). Ritter, Hellmut. Die Volkssprache der syrischen Christen des Ṭūr ʿAbdīn. 5 vols. A, Texte I– III: Band I, Beirut 1967; Band II, Beirut 1969; Band III, Beirut 1971. Wiesbaden: in Kommission bei Franz Steiner Verlag. ––– 1979. Wörterbuch, Wiesbaden/Beirut: in Kommission bei Franz Steiner Verlag. ––– 1990. Grammatik: Pronomen, “sein, vorhanden sein”, Zahlwort, Verbum. Stuttgart: Franz Steiner. Sabar, Yona. 2002. A Jewish Neo-Aramaic Dictionary: Dialects of Amidya, Dihok, Nerwa and Zakho, northwestern Iraq, Wiesbaden: Harrassowitz. Sokoloff, Michael. 2009. A Syriac Lexicon: A Translation from the Latin, Correction, Expansion, and Update of C. Brockelmann’s Lexicon Syriacum. Winona Lake, IN: Eisenbrauns; Piscataway, NJ: Gorgias Press. Tezel, Aziz. 2003. Comparative Etymological Studies in the Western Neo-Syriac (Ṭūrōyo) Lexicon. With Special Reference to Homonyms, Related Words and Borrowings with Cultural Siginification, Studia Semitica Upsaliensia 18. Uppsala. Tezel, Sina. 2011. Arabic Borrowings in Ṣūrayt/Ṭūrōyo within the Framework of Phonological Correspondences: In Comparison with other Semitic Languages. Studia Semitica Upsaliensia 27. Uppsala. Stowasser, Karl and Ani Moukhtar (eds.). 2004. A Dictionary of Syrian Arabic: EnglishArabic. Washington, DC: Georgetown University Press. Thesaurus, see Payne Smith, R. Vocke, Sibylle and Wolfram Waldner. 1982. Der Wortschatz des anatolischen Arabisch. Magister-Arbeit, Universität Erlangen-Nürnberg. von Soden, Wolfram. 1965–1981. Akkadisches Handwörterbuch, I–III, unter Benutzung des lexikalischen Nachlasses von Bruno Meissner (1868–1947). Wiesbaden: Harrassowitz. Wehr, Hans. 1976. A Dictionary of Modern Written Arabic, edited by J. Milton Cowan, 3rd ed. Ithaca, NY: Spoken Language Services, Inc.

 

Amharic and NSM (Natural Semantic Metalanguage) Kjell Magne Yri, University of Oslo

The NSM, also called reductive paraphrase, is the result of a large group of researchers’ project of creating a metalanguage for semantic description that is • • •

non circular built on natural language using elementary elements of meaning that are universal for all languages

The initiator of the project is Canberra-based Anna Wierzbicka, whose Semantic Primitives appeared in English in 1972. The history of the project and the status of more recent theoretical additions to it are summed up in Goddard 2011, which is also the main source of information about NSM including the practical examples used in this article.* The present contribution is an attempt to translate the set of ‘primes’ of the NSM into Amharic, and to use the necessary primes of what I shall call Amharic based semantic metalanguage (ABSM). I shall show glimpses from the translation process in the form of short commentaries, and experiment with an ABSM syntax in applying the terms in the explication of English (be) sad and Amharic azzäna, which roughly may pass for translational equivalents of each other. The elementary elements of meaning are now called ‘primes’, earlier ‘primitives’, and by the year 2011, 63 have been identified (Goddard 2011: 66). As the set is claimed to be based on empirical research, there is a constant discussion of whether or not a meaning is elementary and universal enough to be included as a prime. While the prime itself is the meaning, actual languages may express these meanings by exponents that are either smaller than word level, words, or fixed phrases. The 63 primes are listed here with their English exponents in small caps. They belong to a variety of grammatical, especially semantic categories or domains, which here are specified in parenthesis after the respective *

 

                                                                                                                        I would also like to draw attention to Amberber (2008), who also discusses Amharic translations of the primes. That article is more positive than mine; still, similar points of criticism are put forward, especially highlighting some problems related to polysemy.

570

Kjell Magne Yri

group. The tilde stands for language-specific combinatorial variants (allolexes) and the word following it does not count as a prime. I, YOU, SOMEONE, SOMETHING ~ THING, PEOPLE, BODY (Substantives) KIND, PART (Relational

substantives) THIS, THE SAME, OTHER ~ ELSE (Determiners) ONE, TWO, MUCH~MANY, SOME, ALL (Quantifiers) GOOD, BAD (Evaluators) BIG, SMALL (Descriptors) THINK, KNOW, WANT, FEEL, SEE, HEAR (Mental

predicates)

SAY, WORDS, TRUE (Speech) DO, HAPPEN, MOVE, TOUCH (Actions,

events, movement, contact) (Location, existence,

BE (SOMEWHERE), THERE IS, HAVE, BE (SOMEONE/SOMETHING)

possession, specification) LIVE, DIE (Life and death) WHEN

~

TIME, NOW, BEFORE, AFTER, A LONG TIME, A SHORT TIME, FOR SOME TIME,

MOMENT (Time) WHERE ~ PLACE, HERE, ABOVE, BELOW, FAR, NEAR, SIDE, INSIDE (Space) NOT, MAYBE, CAN, BECAUSE, IF (Logical VERY, MORE (Intensifier,

concepts)

augmentor)

LIKE ~ WAY

(Similarity) (Goddard 2011: 66). The claim of the framework followers is that these meanings have an exponent in all languages, and that the meanings of every single lexical item in any language can be adequately described using these exponents combined according to a simple grammar that is also universal: the NSM syntax. I now try to translate this list into Amharic, playing with the claim that the primes have an equivalent in every language, and keeping in mind that the exponent may be anything from a morph to a fixed phrase, i.e. not necessarily a word. Verbs are given in their customary citation form, the perfective 3SG.M. Some primes translate straightforwardly; others, more problematic, are accompanied by my comments.

 

 

Amharic and NSM (Natural Semantic Metalanguage)

571

I, እኔ YOU, አንተ፣

አንቺ፣ እናንተ

One would suppose that the underlying prime is really the abstract grammatical concept “2nd person”, therefore the three must be counted as one exponent in Amharic. SOMEONE, ሰው SOMETHING ~ THING, ነገር PEOPLE, ሰዎች

SOMEONE and PEOPLE are analyzable in terms of someone + plural. PEOPLE is therefore not needed as a prime, at least not in Amharic.

አካል KIND, አይነት PART ክፍል BODY

THIS, ይህ THE SAME,

አንድ

The numeral አንድ “one” has a remarkable polysemy in Amharic. It is seen to be the equivalent of several purported primes: SAME, ONE, LIKE, and in reduplicated form: SOME. OTHER ~ ELSE ሌላ ONE, አንድ TWO, ሁለት MUCH ~ MANY, ብዙ SOME, አንዳንድ ALL ሁሉ GOOD

ጥሩ

This and the following three so called evaluators and descriptors are also part of the meaning of corresponding comparative verbs, meaning “be better, be worse, be bigger, be smaller”, thus combining with the prime MORE. BAD መጥፎ BIG ትልቅ SMALL ትንሽ THINK አሰበ KNOW አወቀ WANT ፈለገ FEEL

ተሰማ

Several African languages construe the meaning “feel” with words whose primary meaning is “receive some sensory impression”. Amharic employs the passive derivation of “hear”; in Swahili the equivalents of both “see” and “hear” serve to ex-

 

572

Kjell Magne Yri press “feel”. Maybe that does away with the necessity of FEEL as a prime in Amharic.

SEE አየ HEAR ሰማ SAY አለ (alä) WORDS ቃላት TRUE እዉነተኛ DO አደረገ HAPPEN ሆነ

አንቀሳቀሰ TOUCH ዳሰሰ BE (SOMEWHERE) አለ (allä) MOVE

THERE IS አለ (allä) HAVE አለ

(allä inflected for subject (possessed) and indirect object (possessor)) BE (SOMEONE/SOMETHING) ነው

representing all inflected forms of the copula, including the suppletivistic root honthat is also listed as the exponent of HAPPEN. LIVE አለ፣

ኖረ (allä)

DIE ሞተ WHEN ~ TIME ጊዜ NOW አሁን BEFORE በፊት

In Amharic the exponents of BEFORE and AFTER must count as calqued phrases: preposition + noun. AFTER በኋላ A LONG TIME ረጅም

ጊዜ A SHORT TIME አጭር ጊዜ FOR SOME TIME ለጥቂት ጊዜ MOMENT ጊዜ፣

ቅጽበት

WHERE ~ PLACE ቦታ HERE እዚህ ABOVE ላይ BELOW ታች FAR ሩቅ NEAR ቅርብ SIDE ጎን INSIDE ውስጥ

 

 

Amharic and NSM (Natural Semantic Metalanguage)

573

NOT

Negation in Amharic is inflectional with a lot of distributional variants. It can, for example, be described as a circumfix al- ... -mm. MAYBE ምናልባት CAN ቻለ BECAUSE ስለ IF

Expressed by, e.g., bә- and kä- as verbal prefixes. VERY በጣም MORE

በለጠ

(NB: a verb) It would be clumsy to delimit this prime to just one exponent. e.g., bämmibälṭ “to a greater extent” (a calqued prepositional phrase). Amharic has a number of comparative verbs that express that something occurs to an exceeding, equal or inferior degree. LIKE ~ WAY

እንደ፣ መሰለ

The latter is a verb with the meaning “be similar to”.

Let me mention a puzzle: The primes are supposed not to be compositional. “Ultimately, the only way to show that something is not an indefinable element is to succeed in decomposing it” (Goddard 2011: 66-67). How can primes be indefinable if in a certain languages, like in Amharic, the obvious exponents are composed, like AFTER and several others in the list above? I would take that as proof of definability. Apart from my comments in the list there is no room to discuss extensively the choice of Amharic translations of the primes. Let me just mention one: SOMEONE:

How do I find the best equivalent in Amharic? In the following I show how I went about it.

I bear in mind that it is the abstract meaning of SOMEONE that is the real semi-universal prime, not the English word ‘someone’ with its language-specific meaning; that is just an exponent of this prime. Judging from how it is used in the explications in Goddard (2011), I guess that SOMEONE must have an indefinite 3rd person reference, like ‘person’, ‘some person(s)’, ‘any person(s)’, (but I do not know who). Candidates in Amharic: säw ‘man, human’, and säw ‘one person’, mannïññawum (säw) ‘any person’, әntәna ‘a person (whose name I do not recall)’. All may be indefinite as participants in a clause, typically used to introduce a participant that is later referred to as identified, i.e. definite. How do I make my choice? Do I need some background information about how the pro-

 

574

Kjell Magne Yri

ponents of the NSM school have defined the prime (cf. the ‘biography’ of each prime (Goddard 2011: 373)), or can I assume that the qualities of the English exponents are transferrable to the exponents in other languages? For example: the expression ‘this someone’ frequently appears in explications, where ‘this’ is an anaphoric pointer back to a previously mentioned ‘someone’, and thus makes the phrase correspond to a definite referring expression like he or she. Does it mean that the corresponding prime in Amharic should be definable with a demonstrative or the like? As I understand it, the answer is yes, what leaves and säw as the Amharic exponent for the prime (quantified, but indefinite), and ya säw ‘that person’ or yïh säw ‘this person’ as the equivalent of this someone. This is fine for languages where the concept of SOMEONE is expressed by a noun-like word that is modifiable, but not in a language where SOMEONE is expressed by an unmodifiable pronoun. This someone might have to be translated as the equivalent of he/she as the literal equivalent of this someone might be an impermissible collocation in the X language grammar. This is already a problem for a universal NSM syntax. I now assume that I have found the best Amharic exponents of the relevant primes, and want to compare English sad and Amharic azzänä in terms of the NSM/ABSM. An explication of English sad is found in Goddard (2011: 112, based on Wierzbicka 1999: 60-63) and looks like this. The first line puts the word into a context, sets the scene in a way. Someone X was sad at this time Someone X thought like this at this time: ‘I know that something bad happened I don’t want things like this to happen I can’t think like this: I can do something because of this I know that I can’t do anything’ Because of this, this someone felt something bad at this time Like someone can feel when they think like this If I succeed in translating this explication into the ABSM, the result will be a semantic analysis of English sad that an Amharic speaker will readily learn as the meaning of that word in that language. The NSM theory predicts that it will translate easily, but as we shall see the problem of a common MSN syntax is not easy to solve. Obviously English syntax is not equal to Amharic syntax, and if

 

 

Amharic and NSM (Natural Semantic Metalanguage)

575

the explication does not conform to the target language syntax, the result will be pure nonsense. So, here follows my attempt in Amharic, following an Amharic syntax that is as simple as possible: And säw (X) bäzzih gize azzänä (was sad) And säw (X) bäzzih gize әndäzzih bәlo assäbä: ‘mät’fo nägär әndä-täfäs’s’ämä awk’allähu (honä: possible different meaning)

әndäzzih yallu nägäročč әndihonu alfällәgәmm (infinitival complement, also transfer of volition)

әndäzzih massäb alčәlәm: bäzzih mәknәyat and nägär madräg әčәlallähu and nägär madräg әndämmalčәl awk’allähu’ sәläzzih – säwočč әndäzzih bәläw siyassәbu lissämaččäw әndämmičәl – yәh säw bäzzih gize mät’fo nägär täsämma(w) The Amharic version is supposed to be a faithful translation of the English version. If the two words are really equivalent in the two languages, the Amharic version does not only describe the meaning of (be) sad in English, but also the meaning of azzänä in Amharic. That is highly unlikely; probably a lot of different formulations should be included to bring out the difference, but that would need the input from a native Amharic speaker, which I am not, and that detail, important as it is, is therefore ignored in this contribution. Syntactic devises that are necessary in Amharic but not in English become evident e.g. English ‘things like this’ needs to be structured with a relative clause in Amharic: ‘things which are like this’. It is pointed out that ‘this’ can be used to refer backwards to a whole chunk of information, but as soon as it is used this way, it becomes difficult to determine what is the exact intended reference. A trick to help delimit the reference is to put a colon, what is frequently done in the explications. A colon, not part of the metalanguage, helps the reader to think of what follows instead of what precedes. Not much is said about the future NSM syntax in the textbook by Goddard, but it seems to me that a common MSN syntax is impossible. For the explications to make sense at all, they will have to be expressed with the language specific constructions at all levels. In the actual examples throughout the book, the explications follow English syntax, using a number of grammatical devices that are not primes. My translation above, to the extent that a native Amharic spea-

 

576

Kjell Magne Yri

ker can understand it at all, she does so because it follows Amharic grammatical patterns. What does it take for a researcher who is a native speaker of Amharic to master the NSM? She will probably have to learn English, until all the recipes are translated into Amharic. Then she will have to accept bona fide that the team of international researchers have arrived at the best available set of semantic primes, even for her native language. Else she must reject the set as Anglocentric where it counters her own intuition, as in the unnecessary proliferation of primes related to BE SOMEWHERE (from an Amharic speaker’s point of view). This reasoning connects to the arguments for A SHORT TIME, A LONG TIME, and FOR SOME TIME as universal semantic primes. There may be languages that have these meanings as three uncomposed lexical items, but why should the fact force me to think of these word groups (phrasemes) as uncomposed units in English, Norwegian or Amharic? I see no logical reason. This means that there must be a translation manual with the list of primes, authorised by an expert panel, including the biography of each prime and the rationale for including additions. It must be in English at first, and then, when established it should be available online, everything expressed only with the semantic primes and molecules, and a good variety of scripts etc. (The latter concepts are important within the NSM theory, but not treated in this article.) By way of conclusion I would like to encourage the use of the NSM and its language-specific offsprings for semantic description. It forces the semanticist to take meaning difference and similarity, likewise cultural difference and similarity, seriously. Behind my predominantly critical attitude there is the belief that this theory is on to something important, and that efforts to find out exactly what will yield high rewards, for the study of Amharic as well as any other language.

References Amberber, Mengistu. 2008. “Semantic Primes in Amharic.” In: Cliff Godard (ed.). Cross-Linguistic Semantics, 83–119. Amsterdam: John Benjamins. Goddard, Cliff. 2011. Semantic Analysis. A Practical Introduction. Oxford/New York: Oxford University Press. Wierzbicka, Anna. 1972. Semantic Primitives. Frankfurt: Athenäum. Wierzbicka, Anna. 1996. Semantics. Primes and Universals. Oxford/New York: Oxford University Press.